In article , Martin Brown
writes:

The Jeans length is important for star formation, but the stuff which
forms (rocky) planets is only a small fraction of a larger cloud which
collapsed (as described by Jeans) to form a star. There doesn't seem to
be a lower limit on the size of "planets". There is an obvious upper
limit for (gaseous) planets---stars. The sizes of planets are
determined more by accretion, where gravitation is only one factor.

That suggests an interesting question.

Is it possible to compute either by simulation or from observations what
percentage of ordinary matter is tightly bound together (either
gravitationally or electromagnetically) as a function of length scale
(or mass).

At larger scales, dark matter is important, but we don't know what it
is. In particular, we don't know whether it is self-interacting (other
than via gravity) and even if it isn't, it might not be in the form of
isolated particles (though that is what many people assume); the was a
paper by Bernard Carr and co-authors recently which pointed out that
there is still a mass range where it could be in primordial black holes.

At smaller scales, the last I heard, the IMF (initial mass function) for
stars was not computable from first principles. From observations, we
have a pretty good idea what it is locally, but it was probably
different at high redshift.

With certain assumptions, the Press-Schechter formalism allows one to
calculate a mass function, and, not surprisingly (but a good consistency
test and sanity check), this also comes out of simulations with the same
assumptions.