Planetary formation is still not exactly settled. I have analysed as much of the literature that I could get my hands on, and summarised my findings, together with my conclusions, in an ebook, but of course not everyone will agree with my conclusions. First, there is reasonable evidence that Mars was essentially complete within 3 My, and if you accept the hydrodynamic blow-off of a hydrogen/helium atmosphere for the heavy isotope enhancements of krypton and xenon, then the Earth had to be formed before the late accretion disk cleanout. I conclude that must have happened within 1 My. the reason being LkCa 15b, a planet 5 times as massive as Jupiter approximately 3 times further from a star that is estimated to be slightly smaller than the sun, and is only 2 - 3 My old. Accretion must, all other things equal (which they are not) be proportional to mass density squared if stochastic (the probability that two particles in random relative motion will be in the same place). The stochastic Grand Tack model argues planets cannot form further than about 10 A.U. because the matter density is too low. Even if they could, they would be far too slow. My opinion is that the growth is more likely to be basically monarchic, and my model assumes the initial steps, before gravity is predominant, are chemical in nature. The good news is this model predicts the actual differences in chemical composition between the different planets and for the giants, their moons, assuming the moons have the same composition as the planetary cores. The iron isotope data, indicating a very early accretion, appears to be in agreement with this model.
Abiogenesis is a slightly different issue. It should no care when the planet was formed, but it will be critically dependent on exactly what was accreted, and how. In my opinion (and of others) there are two critical requirements: the formation of micelles or vesicles (to act as the start of a cell wall) and of RNA. RNA is critical because only it can reproduce itself and convey accurate information, but not too accurate information, and it has the remarkable ability to catalyse its own hydrolysis if it can get at the 2-hydroxyl of ribose. It catalyses reactions when it can form a loop, and through moving magnesium ions around, it can change what it does (unlike an enzyme) This enables it to reproduce, but also rearrange its order, which means it can evolve. You may note this is what makes certain viruses so annoying - they evolve into new strains if the old one is not making headway. It has to be RNA, because cytosine and uracil are the most likely pyrimidines and adenine particularly is the easiest to make, which is why adenosine is used to solubilise so many cofactors. The phosphate is absolutely required for thermodynamic reasons -to separate the polymer strands when reproducing it is necessary that the strands are soluble in water, which is achieved by the phosphate anion (and the associated cations affect ribozyme catalytic activity) so the problem then is to join the bases to the phosphate. Nature chose to do this through a ribose furanoside, which in turn is quite remarkable at first sight because ribose is the most difficult sugar to make, and while it has about 20% furanose (and the only such sugar with any measurable furanose is solution) it perforce has 80 percent pyranose, so why are pyranose forms not used? My opinion is that an experiment by Ponnemperuma et al way back in the 60s showed the way - AMP and even ATP were made by shining UV light on solutions of adenine, ribose, phosphate and urea (to solubilize the phosphate). The reason lies, in my opinion, in a mechanical transmission of energy to make the phosphate bind specifically at the 5-position. Damer et al have shown that solutions of AMP and UMP in vesicles and exposed to wet/dry cycles on rocks produce something like 100 mer-length polymers in about 4 hours.
So what has this got to do with planetary formation? The planet has to accrete the means of emitting water, ammonia, hydrogen cyanide, and a number of other things. In my opinion, volatiles had to be accreted as solids, such as carbides and nitrides and liberated with water, accreted as the means of joining rocks together during accretion. This, as an aside, is my interpretation of why Venus has such a high D/H ratio: Venus never had as much water because it was hotter there (water can be accreted up to about 500 oC on aluminosilicates) and when making gas, some hydrogen is lost to space. Deuterium is preferentially retained through the chemical isotope effect, which, as an aside, is far more preferential than the usual explanation of photodissociation in the upper atmosphere. The most difficult part appears to be to get ribose, and the most likely route lies in the fact that slightly alkaline solutions of geothermal water contain silicates, which catalyse ribose preferentially. If so, we need a planet like Earth. There will be no life under ice at Europa (no significant nitrogen or carbon, phosphate sinks to the bottom of an ocean because divalent cations make it insoluble, and light cannot get through the ice). Here, theory could save a billion dollars! If that is right, we need a planet like Earth, with felsic continents and water in the habitable zone, which is most likely, according to the above, around G or heavy K stars. M stars probably do not have the heat to make the aluminosilicates necessary for the felsic continents.