The optimal starting point ratios for PQ/ MQ etc have long been published in the standard reference works (see Levenson in the SPSE manual for example), and it seems that an awful lot of the time all that people are doing is inadvertently demonstrating that Kodak et al did the work correctly the first time round.
My own understanding of what people seem to think 'compensation' means is less dense highlights without screwing up midtone gradients via lower average gradient aim. This is essentially what the DIR couplers in XP2 are specificially aimed to deliver, rather than a much longer linear curve at a lower gradient. That people who are messing around with developers often end up doing the latter instead is, I think, the main source of the confusion.
Yes and no. Development inhibition effects potentially allow for better highlight control, finer granularity and better sharpness. Developer formulation is part of the story, but emulsion structure is arguably even more important. It is a pretty complex and (still) rather commercially sensitive area.
Solvency can allow access to (e.g.) iodide placed in the emulsion such that it will produce inhibition effects where it is released; highly dilute metol only (less than 0.5g/l working solution) can produce desirable effects via exhaustion, but not when a source of semiquinone is added; Phenidone and PQ (etc) developers can deliver inhibition effects without needing very dilute single-shot solutions, and the effects can be altered via the P:Q ratio. Adding 1-Phenyl-5-Mercaptotetrazole (PMT) restrainer can produce useful effects allowing a non solvent PQ developer to still produce fine but sharp granularity (and excellent highlight density control, with the added benefit of massive safety margins for those whose process control is poor or more-talked-about-than-enacted). Some of this has been commercialised by Kodak, Ilford etc in the last 30-40 years (and they've outflanked in research many much-talked-up formulae, subjecting them not just to a battery of properly tough tests of sharpness, granularity etc, but also double blind print comparison tests across wide-ranging enlargement sizes).