VE Long Term Fuel Trim vs. Ambient Temp

[10_SS]

I'm looking over my stock log files and I noticed the following Long Term Fuel Trim MAF vs. RPM histogram correlation and wanted to see what everybody thinks.

These are stock for a 2010 Camaro LS3 manual and has very little knock so things are running good:
~45F: ~5% Fuel Added in most cells
~60F: ~5% Fuel added in most cells
~64F: ~6% Fuel added in most cells
~84F: ~3% Fuel added in most cells
~90F: ~0% Fuel added in most cells

Now, I went and updated to a 2 bar enhanced and worked out the SD VE tables pretty good in only the Closed Loop cells, then turned back on the MAF to log that and at:
~91F: ~3-4% added long term fuel trim distributed mostly evenly so I added about 4% to every MAF Airflow vs. Frequency above about 2799hz which lowered the fuel additions to about 0-2%, but then I realized I still had VE tables active so I have to redo it.. but this isnt the point of this post.

The purpose of this post: Are the LTFT swings vs. IAT (ambient air temp) normal and should I target the "stock" Ambient Temp LTFT % error? Since I am going boosted shortly I would prefer to have it rich and pulling fuel rather than lean and adding.. or.. does WOT PE only add fuel based on positive LTFT and not pull it under WOT (if LTFT is pulling fuel)? This would be the only reason I would leave it slightly lean and adding.

[Higgs Boson]

ltft only contributes to pe when adding.