Whoa, whoa, whoa, and whoa.

A didn't say D-23 was compensating. I said it wasn't compensating. The other people in that D-23 thread said it was compensating, and I disagree. Also, Microdol would be more accurately referred to as a D-23 derived formula, not D-25. In any case, these are not compensating developers. If you want a truly compensating developer you need low concentrations of specific developing agents and sulfite - and a relatively strong alkali along with weak buffering. Crawley's FX-2 is an example.

Re how does D-23 work like D-76, it waas designed by Henn to work that way. D-76 does not have 6 or 8 other things in it. The published formula is:

2g/l Metol
100g/l Sulfite
5g/l Hydroquinone
2g/l Borax

Now, consider at the working pH of D-76, Hydroquinone is not directly active, but rather plays a relatively minor role in "regenerating" the Metol. Grant Haist, while working on a solution to the pH instability of stored D-76, showed that by omitting the Hydroquinone and increasing either the Metol or Borax by 0.5g/l, the working characteristics would be indistinguishable from D-76, although capacity may or may not be lower. Hence the formula for Haist's version of D-76:

2.5g/l Metol
100g/l Sulfite
2g/l Borax

A good generalized description of the original D-76 in Haist's book is 2g Metol free base in a 10% Sulfite solution, with just enough Borax added to neutralize the acidity of the Metol salt.

D-23 was originally developed as an alternative. More Metol, no Borax. Since the working pH is lower, development times are different and solvent effect is slightly increased. As a result it is said to produce slightly finer grain than D-76 with approximately 10-20% speed loss, which was determined to be of little consequence in the face of improved consistency, which was of benefit to commercial labs.

The point is, you can end up with virtually identical results from different formulas. Just because one formula contains more compounds doesn't inherently mean anything without looking at the interactions between those compounds. Further, the type of film is an important variable.

D-25 was a lower pH version of D-23 (by adding bisulfite) for extra fine grain, since DK-20 was too solvent for more modern films at the time. Microdol was an evolution of D-23 with the addition of Sodium Chloride, which seemed to produce the extra fine grain effect of D-25 but with less chance of provoking the formation of dichroic fog with evolving films. The formula is commonly assumed to be 5g/l Metol, 100g/l Sulfite, ~30g/l Sodium Chloride. However the actual formula has never been disclosed. Nor has the formula for Microdol-X, which included at least one extra compound - an anti-plating/anti-sludging agent.

Pyro is a different subject. Assuming we're talking about staining developers, there are different formulas that produce different results (and Pyrocat HD is a Catechol staining developer, not Pyrogallol). It is also important to note it can be the colour/intensity of the stain itself which produces compensating characteristics on certain printing papers, not the shape of the silver curve. Experimentation is required.

I'm not sure what you're trying to achieve. That might help you choose the right developer. If this film you are using is a high contrast emulsion, then a compensating developer, or a special low contrast developer might help. Otherwise you simply dilute your D-76 or D-23, try to reduce agitation somewhat, and develop to a low gradient. You will lose film speed, but that may or may not pose a problem.