Dear Elyrajo,

For a long answer, please read the 1st part of Multimedia File Viewing and Clickable Links are available for Registered Members only!!  You need to Login

The short answer is that synRate is the product of the synonymous substitution rate and the branch length (they are confounded, this is a standard shortcoming of all time reversible substitution models).

Removing synRate has the effect of fixing the length of every tree branch at some arbitrary value (hence a worse likelihood), and making it global makes all branches have the same length, but now estimated from the data (better than not having it at all, but worse than estimating each branch length individually as should be done).

Cheers,
Sergei