FWIW, I have a 408 stroker, with a .544 236 hydraulic roller cam, 9.9:1 comp, Edel Heads, and run a Carter 600 on the street, and a Holley 750 DP (72/84), only 2 corner idle, on the strip into a M1 single plane airgap, with Hooker comp 1 5/8" headers. I bought my Edel heads used, and had them gone through, reground, new seals, new springs, shimmed, etc, before I had them installed. I have a FBO distributor set up with 12 degrees timing at idle, and 32 degrees all in by 2900 rpm. (I want to try advancing it a bit at the strip this weekend). I have a vacuum distributor, putting in another 12 degrees at high vacuum. Everyone told me to run off of the manifold vacuum, which I did for a while, and experienced the exact problems you are experiencing. Because vacuum was coming from the manifold (throttle body) vacuum, I was getting around 20-24 degrees timing at idle, and it idled LOUSY! Vacuum fluctuated from 5 to 10 inches, engine was surging as much as 300 rpm, timing would jump around 2-3 degrees using a timing light on the damper. The engine would even spit and die occasionally, not to mention the occasional backfire.
On a whim, I switched to the idle circuit vacuum (on the Holley metering plate) for the distributor vacuum, and the aforementioned symptoms disappeared. I still have the vacuum kick in, but it is over 1200 rpm with partial throttle. Driveability went up, idle speed smoothed out big time. It still has some erratic vaccuum reading, and the idle speed does fluctuate some 50 rpm or so, but it will idle at 700 rpm, and runs MUCH better! Big cams are fun, but they will idle rough, which is kinda the point, eh?
So, my timing is now at 12 degrees, distributor vacuum unplugged. If I replug the distributor vaccuum into the Idle Circuit, the timing does not change at idle. Before, when I had it plugged into the manifold vacuum, it was adding another 10 degrees, which killed the idle performance!
Jim