And remember, no aluminum Piper ever built, and damn few Cessnas ever
got alodined.

This is not *exactly* correct. Every zinc chromate painted part in a
Cessna (or Beech or Piper) received iridite before painting. Paint
won't stick right without it. Part of MIL-C-5540 or whatever (I'm
rusty).

If you ordered the hideously expensive "corrosion resistance" package,
you got an extra couple hundred pounds of paint -- but no increase in
useful payload. Guess how popular THAT was?? The oil filter adapter
was treated and then clearcoated leaving it golden color (and priced
accordingly).

You don't mention how "clean" your part is. Are you taking off the
Clad coating? If so, the parts really need to remain *submerged* until
coating. Any exposure to air will instantly start oxidation and reduce
coverage quality (usually judged, as you're doing, by depth of color).

My theory is that your initial coating, utilitzing your stated process,
is merely adhering to an oxidation layer that, of course, scrubs right
off. The most of the surface was "sealed", if you will, against
further oxidation and once re-treated, responded in a more predictable
and acceptable manner. Just a guess.

Larry