"But it does not necessarily follow that because the placebo (end of study) vs ABL (end of study) comparison was significant, so also was the ABL (end of study) vs ABL (baseline) comparison."

You're right, because it depends on the degree of variation within the groups being compared.  We need the standard deviations or confidence intervals for the group values (and in particular for values at baseline and end of study) to determine if the 4 pt improvement in MOCA in the treatment arm is signficant.  I would be surprised if it wasn't, though, as RVX thought it had the power to detect a change in MOCA of about 2.5 for these sample sizes (make that 2 - see below):

MoCA score change from preliminary blinded data shows a standard deviation of 3.2 points and a sample size of 54 subjects per arm to provide a 90% power to detect a mean between group difference of 2 points at p<0.05."

Jupe