Something like continuous agitation, e.g., rotary processing, would slow down or stop such diffusion by replenishing those areas with fresh developer and thus significantly slow or stop altogether the formation of edge effects. Reduced agitation would have the opposite effect (as seems true in my experience).
If agitation is above the level required to produce truly uniform development, then the effective edge sharpness differences between intermittent and continuous agitation are essentially nil. This has been investigated very, very thoroughly across the industry using microdensitometry - and Richard Henry also came to the same conclusions from his microdensitometric testing. Anyone without real microdensitometric data who is claiming differently is probably reporting other phenomena (lower overall density, printing on harder grades, etc, etc) that by other means are effectively increasing apparent edge sharpness, but it is not a direct result of the developer formulation. I think people need to understand in the clearest possible terms that most of what are claimed as 'innovative' developers and agitation techniques by various garden shed tinkerers can in fact be found being comprehensively outflanked in industry publications, patents & theses, where upon subjection to meaningfully rigorous testing and double blind print comparisons, they were found wanting vis-a-vis D-76 or similar. Huge amounts of time and effort were expended by extremely able chemists and engineers across the industry for decades to try and best D-76 in front of double-blind print panels. Well controlled development in D-76 with continuous agitation is quite capable of producing results that would easily fool the supplicants of various staining developers (i.e falsifiability) - the results of many claims about staining developers seem, in my experience of the people who work with them - and the results attained, from those developers often effectively running out of steam in ways that effectively compensate for a lack of understanding or enforcement/ reinforcement of basic process control - which more flexible and effective mainstream developers show up with rather unyielding and brutal clarity, especially on materials where the manufacturers took the risk of taking the training wheels off from the 1980s onwards.
Also, it would seem to me that developers that contain small enough amounts of sulfite so as to be non-solvent developers would inhibit general inhibition of development in dense areas simply because a smaller amount of developer-inhibiting compounds are being produced and, with frequent agitation, also the formation of edge effects.
Solvency is important to release inhibiting agents from the emulsion, but as I outlined, some developing agents can also produce inhibition effects under specific circumstances as well. So yes, you can produce inhibition effects from non-solvent developers, but again because of the way the mechanism seems to work, as long as agitation is over the threshold of even development, it has no meaningful impact on anything apart from density (which itself does have an impact on effective edge sharpness) which can be controlled for.
Only in very specific circumstances (e.g. halftones on litho) might there be a need for a period of intense agitation (while the chemistry diffuses into the emulsion) followed by absolute standstill - but that relates to the characteristics of litho development and halftone formation.
With infrequent agitation, however, non-solvent developer, then might deliver more pronounced edge effects, since dense areas are less inhibited and develop more fully while the inhibiting by-products that are produced have the most influence on the lines of interface between denser and less-dense areas.
In the kindest possible way, no. What seems to have excited manufacturing R&D about PQ was that it could produce useful inhibition effects in a way that was controllable (relatively at will) and largely insensitive to agitation (if over the threshold of evenness) while allowing replenished systems to be designed in ways that were seemingly not possible with developers that exploited Metol's exhaustion effects.
