I agree with you and others in principle. The only exception I see to this is to keep the blind on any potential improved manufacturing version being tested as equivalent by site(s) and not by year.
You said you seem to think that the last 31 have a higher concentration of methylated MGMT promoter. Can you show me again where you saw that as I missed it the first go round. By the way, IF this group was selected for methylation might it not be to improve statistical analysis of the methylated group and thus important to keep blinded? I know your other points and evidence about this group as you made that clear. Best wishes.