Mars Yang, first just to clarify, I've never tried iodoTMT based enrichment, but worked with TMTs and LFQ quite a lot, so take my reply with a grain of salt. ;)
My main question is: what is the sample that You start with? Is it possible that proteins are reduced at the start (e.g. buffer containing DTT is used for storage)? If yes, You alkylate all of the cysteines with IAA in a 1st step, so in 4th step You do not really introduce iodoTMT (and then loose all of the peptides during anti-TMT enrichemnt). To test this I would try to run sample after step 7 and check if You can see any TMT-labeled peptides.
Regarding general tips - when doing general proteomics (so aiming to have as many quantifiable peptides as possible), think about everything that comes into contact with Your protein/peptide sample. Use only low protein binding plastics (epps, pipette tips, etc.) from the start of the protocol. Do not put your sample into glass containers (use dedicated high-recovery LC vials). Try to validate every step that has immediate impact on recovery (C18 desalting, precipitation) using standards.