Tim,
As you know, sodium ascorbate is just a salt of ascorbic acid. Some developers call for one or the other, but they function the same once the pH has been adjusted to the correct target level. The active developing agent is ascorbate ion -- how it gets into the solution is less important. Ascorbic acid tends to be more stable in the solid form and is generally easier to get, which is why many recipes call for it. These formulations call for more base to compensate.
(I would try to avoid thinking of "accelerators" in a developer. There are just too many misleading connotations. The principle is that developers need to be alkaline to function, and pH has a huge effect on activity, so we use various bases -- borax, metaborate, carbonates, sodium hydroxide, and as discussed below, triethanolamine -- to get the solution to the correct pH.)
Triethanolamine, as Pat mentions, is a base that can also serve as a viscous organic solvent. Ascorbic acid and phenidone (and its cousins) can be dissolved in it, and these organic solutions are protected from the oxidation that usually occurs in aqueous solutions. When diluted into water, the TEA loses its solvent role and simply brings the developer to an appropriate working pH.
Propylene glycol and ethylene glycol are also viscous organic solvents, but they are not bases. In one version of my "Instant MYTOL" formulation (posted here on APUG and on my website) I call for a solution of ascorbic acid and phenidone in a mixture of TEA and propylene glycol. This mixture contains just enough TEA to bring the final developer (once diluted) to the correct pH, but because that amount of TEA isn't enough to dissolve all of the ascorbic acid and phenidone, I added some propylene glycol to get everything into solution.
TEA is in some ways an ideal base for use in B&W developers. The negative-log acidity constant of its conjugate acid (pKa in chemical terms) is around 8, which is also the target pH for ascorbate developers like Xtol. Kodalk (sodium metaborate) is comparable and also a good choice.
Phenidone and dimezone-s are closely related. Dimezone-s has a somewhat different substitution pattern (a couple of methyl groups) that make it more stable to aerial oxidation but on a per-molecule basis, I think their activity is pretty close. From what I recall, dimezone-s was primarily invented to get around the patents covering phenidone. Neither are acidic or basic. Metol and hydroquinone are very different beasts and interact differently with other developing agents.
Sodium sulfite plays many roles in a developer, but from the point of view of formulation design, it's important to remember that it itself is somewhat alkaline and will raise the pH of solutions. In some developers (D23 and the Microdol / Perceptol types) sodium sulfite is the only base present.
Boric acid is sometimes used in combination with metaborate or borax to adjust pH. Salicylic acid is a chelating agent for metal salts that some ascorbate developers use to prevent "sudden death" by metal-catalyzed oxidation.
I hope this information sets you on the right path. I think getting a handle on acids / bases, pH, what happens when substances are dissolved in water, etc. would really help. Without this understanding, you may find developer formulation pretty frustrating.