Funnily enough, cosmology's use of "baryon" predates the Standard model (it even predates Politzer, Gross & Wilczek 1973) and leans heavily on the Greek root "b

Funnily enough, cosmology's use of "baryon" predates the Standard model (it even predates Politzer, Gross & Wilczek 1973) and leans heavily on the Greek root "barus", meaning heavy i.e. having a significant rest mass (and thus implying "nonrelativistic", more below). Since then we've developed the Standard Model and its baryons have many properties that are essentially irrelevant at cosmological scales after reionization. More technically (and all to the first order), in \Lambda-CDM the first Friedmann equation can be written as H(a) \equiv H_0 {\sqrt {(\Omega_c + \Omega_b) a^{-3} + (\Omega_{rad+hdm}) a^{-4} + ... }}, the \Omega_{...} being density parameters. Although in this form "c" stands for Cold Dark Matter (CDM) and "b" stands for baryons, the important common attribute here is that the rest mass is high enough that absent highly atypical kicks they move slowly compared to photons[1]. I explicitly added "+hdm" (hot dark matter) to the radiation term, where hdm is mainly relativistic neutrinos, which aren't heavy and thus usually move at speeds closely approaching that of photons. The relativsitic/nonrelativistic split is important in structure formation, roughly because the former do not stick around during gravitational collapse. One can consider here that \Omega_c is (still) likely to be mainly nonrelativistic heavy uncharged leptons and that if we lump nonrelativistic charged leptons separately into \Omega_{b+c+l} term they would be a tiny contribution to the density. Relativistic baryons and relativistic charged leptons could be lumped in with \Omega_{rad}; it wouldn't make much difference. Experimental values for \Omega_{rad} are very small (~ 10^-4 vs \Omega_{b+c} ~ 3 x 10^-1, i.e. about 1:3000) and are often dropped for ease of calculation. Bound states are either relativistic or not, such things would just take the place of relativistic neutrinos and photons from \Omega_{rad}. Doing any of this is just fiddling at the margins, however. Cold Dark Matter and nuclei simply dwarf any other matter component across most of the universe's history. - -- [1] In the FLRW model the components of the \Omega_{...} densities (i.e., matter in the large) are homogeneous, compressible fluids of constant density at rest under a constant spatially isotropic tension in an "equatorial" slicing of the RW spacetime, so we don't even have relativistic baryons (for example): everything is moving inertially as spontaneous symmetry breaking freezes them out of the hotter denser earlier phase closer to the big bang. This is a reasonable approximation at the largest scales.