From what I understand, this isn't quite correct. The rest of the universe will essentially disappear from us as you state with the exception of the local group: the nearby galaxies that our galaxy is gravitationally bound with. And if I understand the expansion of the universe correctly, it will not effect gravitationally bound objects unless of course the expansion accelerates to the point to exceed that gravitational bond.
Well, that's one possibility, however, I think people are confusing two (technically actually three) different ends of the universe here. Heat death, as initially asked about, does not involve any of the universe disappearing from view. If the expansion of the universe is constant or accelerates at a constant rate, or if the universe is open or flat, then you get what is properly called the "Big Freeze", which is only one out of several possible heat deaths. The expansion would never overcome gravity or cause the rest of the universe to disappear, things would simply become very cold and very boring.
Heat death can also refer to a few other possible scenarios, all of which involve the universe reaching a state of maximum entropy. In a flat universe with no acceleration, the universe tends to a finite size after an infinite time. This means that although it will never cool down as far as in a Big Freeze, everything will be in equilibrium with everything else and so there would be no work possible and nothing would ever happen. Another heat death is also possible in a Big Crunch, where the universe stops expanding and contracts down to a point. Depending on the exact way this happens, the universe could reach a thermal equilibrium before the final collapse and undergo a very hot heat death.
The only scenario that involves bits of the universe disappearing from view is the Big Rip. This occurs if not only the expansion of the universe accelerates, but if the rate of acceleration is also increasing. However, in this case gravitationaly bound objects will definitiely be ripped apart as well, it will just take a little longer. Eventually, everything would become seperated from everything else, since you would need to travel faster than light to get anywhere. However, things wouldn't disappear entirely. What happens at an event horizon is that time appears to stop to an external observer and redshift becomes infinite. This means that instead of things just disappearing over the cosmological horizon, they would appear to slow down as they approached it, and get redder and fainter. Eventually you just get a smeared out faint blur. Opinion is still divided on whether there is any information contained in the radiation from an event horizon, but I believe the general concensus is currently that there is.
Actually, by the time of the Heat Death of the Universe, there will be no light, no galaxies/stars/dust etc. All mass will have decayed/turned to it's constituent energy and that energy, in the form of heat will be even throughout the universe - and the whole universe will be at a temperature a few billionths of a Kelvin over absolute zero.
Sort of. In addition to the above, the part about all mass decaying probably isn't correct. Leptons like electrons and neutrinos don't decay, so there will always be plenty of them around. When people say that matter will all decay, what they mean is that protons will. However, this is far from certain. Some theories suggest that protons decay, but have an insanely long half-life, however, there is currently no evidence that they actually do. Experiments which have tried to observe proton decay have determined that the half-life msut be at least 10
35 years, and I think theory suggests it could be 10
36 or even longer.
Proton decay is actually very interesting, since it cannot occur in the standard model and is therefore one of the few ways of testing theories which try to replace it. Although given the long half-life it's not a very easy test.
