Interesting result: When developing a complete roll, you would expect a developer to develop less (make thinner neg's) than with a test-strip. That's because with a whole roll, there's much less developer present per square cm of film, so it gets more exhausted. So why did this happen with the baseline developer?:
With the roll, the developer became more active! It overdeveloped the roll, using the same time and temperature (and same pH). I'll guess that the by-products of the phenidone and AA accelerate development. I remember reading about this, but this is the first time I've seen it. I'll reduce development-time a little, unless somebody has a different idea...