I finished documenting the majority of what I have so here are the variations that are below 5% when taking the backs into consideration.
For comparison the Drabek white borders is 1.754% and the Bush no code is 2.439%.
The two that have an asterisk next to them only have one example so I will have to check and make sure I didn't make a mistake on them.
36 Missing parts of company lettering bold logo no glow 4.444%
80 Led league in earned runs in 1990 no glow no bold logo 3.508%
100 101 hits in 1990 no glow no bold logo 3.773%
*167 Harrisburg correct no glow 2.222%
249 Missing parts of company lettering bold logo no glow 2.50%
*277 No pink feather in cap glow 1.923
324 7 home runs no glow no bold logo 4.081%
378 Has 1990 Port Charlotte & Birmingham stat lines no glow 3.921%
454 has Omaha stat line no glow 4.166%
527 #105 Kevin McReynolds no glow 3.921%
599 Has Syracuse stat line no glow 3.571%
687 4.46 ERA in 1990 no glow 4.545% (I'm surprised at this one)
I went back and checked and I do have
one #277 Scott Coolbaugh no "pink feather" in cap with a glow back.
The "pink feather" was never partially or fully removed it's just a matter of which back the card has if it's a glow back it will have the "pink feather".
All of the partial "pink feather" examples I have came from the very early print runs with the A* B* sheet codes or the Trebelhorn A* code cards
and the full feather is still found in the very late print runs with all of the corrections made.
I still haven't found the one Tomlin example but I haven't checked all my cards yet.
img644.jpg
I also checked my Comstock cards because I thought the one Cub variation with a non glow back was a mistake on my part so
I didn't initially even list it but I do have only one example so that is another variation that falls below the 5% at 2.702%.
img644 - Copy.jpg