Reply
Any scattering mechanism (if significant) should lead to a broadening of features, both spatially as well as spectrally (even for 'elastic' scattering you would get frequency shifts due to the Doppler effect). However, for any reasonable value for the intergalactic electron density, the scattering probabilities would be anyway much smaller than 1 (even over distance like 10¹⁰ lightyears), especially as the photon energy could only be reduced by a very small amount per scattering event (see the last equation on my page regarding the Photoelectric Effect (in the 'Particle Model' paragraph)). On the other hand, if one hypothetically assumes enough scattering particles to be present, what would happen is that the original feature would be surrounded by a 'halo' of scattered light (similar to the moon showing a halo when observed through a thin cloud layer) and it is only the light of this halo that would be redshifted, but not the direct light from the object (which would be merely reduced in its intensity). If the optical depth of the scatterers is much larger than 1 (which it would have to be if the Compton effect should lead to substantial changes of the wavelength), one would in fact not see the original object anymore at all, but just a diffuse background of scattered light.

But anyway, as mentioned on my home page entry regarding the Compton Effect , the usual interpretation of this phenomenon is flawed in my opinion i.e. it is actually not a scattering process at all, even for the standard experiments made with x-rays and solid targets (I am for instance not aware of any experimental confirmation of the Compton effect for actually free electrons i.e. in a plasma). Also the concepts of energy and momentum conservation should actually not be applied to light (see my corresponding pages regarding 'Photons' and Energy and Momentum Conservation).

Vlad Tarko (2)
What is the "reasonable value for the intergalactic electron density"?
Grote Reber, who made the hectometre radio telescope and detected radiation coming from the apparent void between the galaxies, said that based on his observation, and if one assumes that the radio waves are emitted by plasma existing between the galaxies, one can compute the amount of plasma (protons and electrons) existing between the galaxies. He said that more that 99% of all the matter in the universe in is this form!
So, maybe the light coming from galaxies and passing through this intergalactic plasma is faced with significant Compton effect (see for example "Compton Effect Interpretation of Solar Red Shift" by J.W. Kierein and B.M. Sharp (they also present a different interpretation of what a quasar is: just a star with a large atmosphere, not so distant from us after all)).

Reply (2)
I am actually aware of the theories of John Kierein and others, but as I said before, I can not accept this as a possible explanation as a) one would need intergalactic electron densities of the order of at least 10² cm^-3 for a redshift of about z=1 (which would be about the plasma density of the solar wind or interstellar HII-regions), b) even if one tweaks the Compton scattering theory so that the effect becomes more efficient (as Kierein does I think), the obvious effects of the scattering mentioned above could not be avoided, and c) in my opinion the Compton effect doesn't exist anyway for free electrons.

However, scattering is not the only process by which matter can affect light. For instance, a magnetic field can rotate the polarization vector of light. In this sense, I have suggested on my website plasmaphysics.org.uk under Plasma Theory of Hubble Redshift of Galaxies page that the irregular electric field due to the electrons and ions in intergalactic space causes the redshift. Even though this is a statistical effect as well, this is very much different from scattering but could probably be compared to a refraction effect (albeit one independent of wavelength): the point is that here the effect of the plasma field in the direction of propagation of the wave always has the same sign, that is the stretching of the wave (and thus the redshift) is additive (i.e. it is a scalar effect) and will, despite the random nature of the field, result in a very sharply defined redshift.
On the other hand, the transverse deviation of the direction of propagation caused by the plasma field has vector properties and thus a random walk character given the isotropy of the medium. So there will be some blurring due to this circumstance, but this should be very small and negligible: first of all, if one assumes the intergalactic plasma field to have a scale of about 1m (corresponding to a plasma density of 1m^-3) then, over a distance of 10¹⁰ lightyears = 10²⁶m, one has 10²⁶ additive redshifts. Now this leads to a redshift of the order of z=1, i.e. in the space of 1m the redshift change is about dz/ds=10^-26/m. Assuming that the direction of propagation is changed by the same amount (compared to 360 deg) within 1m, this results over the total distance of 10¹⁰ lightyears in a statistical angle of deviation (i.e. a blurring) of Δα=10^-26.√10^{26 .}360 deg = 4^.10^-11 deg, which is negligibly small (for comparison, the angular width of our own galaxy from a distance of 10¹⁰ light years would be about 6^.10^-4 deg , i.e. about 7 orders of magnitude larger; it would take a distance of 7^.10¹⁴ lightyears until the blurring would become comparable to the apparent size of the galaxy (which would then only be about 10^-8 deg; of course, this would be not observable anymore as it is beyond the 'horizon' at about z=10⁴ caused by the spatial scale of the intergalactic plasma field (see the page Plasma Theory of Hubble Redshift of Galaxies at at my site plasmaphysics.org.uk).
Furthermore, as indicated on my page regarding Galactic Redshifts and Supernova Lightcurves, the apparent delay in the light curves of supernovae could also be explained by this effect.

Regarding quasars: it is possible that the redshift is caused by their atmosphere, but again this cannot be due to the Compton effect (for similar reasons as above). In my opinion it could be caused by a strong electric field surrounding the object (e.g. a plasma polarization field associated with strong plasma density gradients). I am not sure however if the very strong electric fields required could be produced in this way, so may be the redshift of quasars is indeed due to the same mechanism as for normal galaxies, and that thus their intensity is really so enormous.

Reply
I do largely agree with your remarks, but so would actually in this case also quite a few 'mainstream scientists' (at least as far as the evidence for black holes is concerned). First of all, the presence of a large central mass in the centre of galaxies (or at least in our own) is obviously unquestionable as proven by the observations of stars very close to the galactic center. However, although there is no direct visible evidence of such a large central mass, this does of course not prove the existence of 'black holes' in the proposed theoretical sense that it is an object with an 'event horizon' (i.e. from which no information can escape). There have been claims based on the observations of other objects (X-ray binaries) that the 'event horizon' would have been 'detected' given the unexpected dimness of the observed radiation (see this NASA article) but this claim is very much contested by other scientists (see for instance this reference) who indeed claim that a direct proof of the 'event horizon' by means of observation in the electromagnetic spectrum is fundamentally impossible.
I can actually suggest my own theoretical explanation of the dimness of the radiation from these massive objects here: in gravitational equilibrium, their gravitational energy and thus their temperature increases proportionally to the mass and inversely proportional to the radius (see the page regarding Coronal Heating on my site plasmaphysics.org.uk); so one would expect extremely high temperatures for these objects (a supermassive object of 1 million solar masses and with solar density would have a temperature of about 10¹¹K, corresponding to particle energies of 10⁷eV), and the point is that the collision cross sections of all radiative processes decrease very strongly with energy E (between E^-2-E^-3; see for instance my page regarding the Radiative Recombination Cross Section), which simply means that such a hot object becomes virtually invisible as it only emits a very small amount of radiation at very high frequencies (in the gamma ray region). This could explain the absence of any observed radiation from such massive objects without involving black holes at all.
Regarding the alleged effect of masses on light in the form of 'gravitational lensing', you may want to have a look at my page Plasma Theory of 'Gravitational Lensing' of Light which suggests that this is also a plasma effect rather than directly related to gravity.

Reply
The point is that at the moment when the light of the other galaxy was emitted, we would have been much closer to it than we are now if we would have been receding from it since then. So if we had just a distance of 0.7 billion years at that time, then, in view of the constancy of the speed of light (the travel time of the light signal should be independent of the recession velocity of the light source), we should thus have observed the corresponding redshift already 0.7 billion years later i.e. 12.3 billion years ago, but not now. So if the universe was expanding, we would not be able to look back so far into the past in the first place.

If all the galaxies had been created at about the same time, one should also generally see galaxies with high redshift in a state very different from nearby galaxies. However, there is no conclusive observational evidence for this. The universe at high redshift looks pretty much the same as that at low redshift (apart from the fact that it has a larger redshift (and maybe some biases associated with this)).

For an alternative redshift theory see my page Plasma Theory of Hubble Redshift of Galaxies on my other site plasmaphysics.org.uk.