I thought the vacuum signals used by the retard and advance are different. For the retard it is hooked up to the manifold, and thus gets a strong vacuum signal at idle. But the proper signal for the advance was on the carb body in the venturi, before the throttle - which means it only gets a signal when there is a lot of air going through the carb - ie WOT. I could most definately be wrong on that detail though.

I recently took apart my Mallory distributor's centrifugal advance mechanism, cleaned, greased and reassembled. I think it had been sticky before - the timing was doing funny things. Worked much better afterwards, and I tweaked the springs around a little and finally got rid of the either/or timing scenario I previously had (either good low end torque or good high end power, not much middle ground). I'm starting to think more seriously a modern fuel injection system that also has computer controlled timing included. I think tinkering with numerical tables would be a bit more direct and less arcane than trying to approximate the proper curve with some springs and weights.
--
I'm JohnMc, and I approved this message.