Jay, one problem with wrong buffering agent is inconsistent developer activity across batches. Because the solution pH is not buffered, the solution pH is very sensitive to small measurement errors when mixing each batch. The same inconsistency problem occurs when the developer is aged. It's very well known from early 20th century. Because of this, D-76 formula, for example was compared against several of buffered versions. Of course, a buffered version is superior in consistency, so that's what's used today.
Developers for practical fine-art and commercial uses should be made with robustness as a design spec, whenever possible. Ad-hoc approach may produce acceptable result sometimes, and that may be good enough for casual chemistry experiments for non-serious applications, but it is the consistency and reliability that matters a lot in production use.
A big part of the problem can be instantly removed by replacing the buffer with the correct agent. Why you insist in not using the right agent? I don't see a good reason.
Regarding your provicative comment, I'm not really concerned about your work or Gainer's work being a competition for my developers. If that were the case, why I don't have issues with Sandy King? I don't know him personally, but the way he writes and reports his work is pretty fact-oriented. He also studies similar work previously done by others before jumping the gun to claim his work novel.
What I'm concerned about is that a lot of unsubstantiated claims (and stories) disseminated frequently here, photo.net, and probably other places.