I recently replaced my smooth idlers with toothed idlers in hopes of reducing VFAs (after seeing from positive results from 2 other people). But, there was no visible change to my standard test print:
I also found about 0.003" TIR on the B-motor pulley. But, that’s less than 1% variation on the effective circumference.
I’m very curious if a more rigid, dimensionally pure, and linearly constrained X axis nets you an improvement. Keep us posted!