ah...stats count are never correct...;-) ...ignore them....If I remember correctly they are counted without taking the merge mode into account....so a missing parent rom may be listed only once in the scan tree but multiple times in the stats....etc..etc....really...the only important thing is a clean tree output.
Regarding 317-0432-jpn.data..let's look at 187...there was it only available in "vs42006". Now with 188, "vs42006b" was added which uses an alternative chd. So as long as you don't have the chd, you're actually missing the set "vs42006b"...and if the set is missing, it will list all its roms as missing.....even if in this case it belongs to the parent.
When you've added the chd, the set check found vs42006b as available and the rom was not listed as missing...
Nothing that special....and I think I've changed it recently in the nightly build...but I will double check that.
And the pippin set was a bit special this time anyway since an existing file was doubled for whatever reason....so rebuilding the set itself would fix the problem.
Update:
ok...I know the reason why cmpro did is not able to 'fix-missing' Pippin....it simply skips the found match in its own set....Guess that can be changed for the next version.
Regarding 317-0432-jpn.data, that issue should actually be fixed in latest official cmpro (fixed: don't list existing parent roms in completely missing clones' output) and I cannot reproduce it
Stats count in general could be better with the latest nightly build