I'll start with a quote from Sherlock Holmes "when you have eliminated the impossible, whatever remains, however improbable, must be the truth?". It is hard to know what is impossible and possible with regards to comets. The original pristine born from a molecular cloud hypothesis is fairly solid as a starting point. However, the question of "why are they still here?" appears to be the most pertinent. By all accounts, orbiting bodies essentially are finding a minimum energy, and the tendency is to coalesce into planets and moons in the general plane of the ecliptic. The expectation is that they shouldn't be able to persist over billions of years moving from the Oort cloud where they formed, to the scattered disk or the Kuiper belt, cross by Jupiter without hitting it, then go into semi-stable but often highly elliptic orbits. Creation scientists have jumped on this lack of the actuality emanating directly from the solid physics that explains the planets and asteroids very well but not comets, by claiming they could only have been placed there supernaturally (Me and Sherlock would rule this out as impossible).
Proximally, the comets are following the laws of gravity fairly precisely, barring a few exceptions of non-gravity accelerations. Tracing back and forward, it is often hard to predict past certain points where they invariably go past large planets. With robotic spacecraft, these are called gravity assists, and they save a lot of power, but require very precise calculations because a small error can send the craft into the planet or shuttling off in a helpless direction. Therefore, although the orbits, and even the breakup into families etc. is individually explainable, the overall picture of comets appears very contrived, trajectory wise. From the semi stable orbits they are in (Oort Cloud, Scattered disk, Kuiper belt, Jupiter group, Sungrazer), random interactions, perturbations or collisions would be *expected* to send them on a collision course, or to a trajectory from whence they cannot get back to one of the semi stable orbits possible.
With the discovery of CAI's from comet's coma as captured by stardust in 2006, the prospect that radioactive aluminum with a half life of half a million years was likely to have been intercepted by pristine comets, have increased the expectation that a fair percentage of comets had liquid three dimensional lakes for at least millions of years. I don't think it is a stretch to surmise that since CAIs are almost certain to be directly from a supernova remnant dust cloud, and proximally from the accretion disk of the forming sun, that heavy elements could also have made their way from the same general source to the comets or molecular clouds.
The dirty snow simulations done in the lab generate aggregate that has a density of about .2. Comets with known density have a density closer to .6, so although the outside coating may be low density fluffy dirty ice, the centre is more likely to have a liquid, or formerly liquid, more dense material.
Comet 103P/Hartley gives me an impression that it is made up of two sections with a smooth neck in between. It would seem plausible to me that the comet will eventually break up into two comets, both of which have a very dark exterior with subsurface ice and certain gaps in the dark exterior. Could it be classed as reproduction? Several other cases of relationships between comets with quite different current orbits can be traced back to a common speed, trajectory, and point in space and time.
Assuming a time that comets had liquid water, autocatalytic cycles of the kind where various autocatalytic species would be competing for substrate would occur within the liquid portion, protected from ionizing radiation. The outer snow hull would serve to insulate from ionizing radiation, and also to protect from more minor impacts. These impacts would not directly affect or stop any delicate chemistry, but would supply more substrate. Hot dense materials intercepted would work their way to the gravitational centre, while low density matter may add to the bulk of the protective shell. Freezing and melting cycles (assuming elliptic orbits, solar heating would vary dramatically between perihelion and apehelion) would be a natural form of chromatography - separating different organic chemicals. Nuclear particles from beta decay would also have an affect on the chemistry.
It would be an expectation that the chemistry would continue to evolve, and not have a limit of say an RNA world or single cell organisms. DNA could coexist fairly loosely with RNA, single cell organisms and various bases, enzymes, proteins etc. Chemical evolution and biological evolution could happily coexist while the centre stayed liquid. If Earth is considered a perfect environment for higher animals to evolve, the inside of comets should be even more perfect. It wouldn't be just a possibility that evolution could go on to well past the sophistication of Earth's evolution, but an actual *expectation*.