Are you running ported vacuum to the advance?

How about the transition as you accelerate from stop? Smooth or hesitation in any way? The bird tranny always starts in second and it stumbled a bit for a while until I got the primary float level to where it liked it. I just had it a little too far on the lean side - which will cause the odd misfire too.

I've got an Accel "HEI" four post operating module that I've done some tests on. HEI units don't use a ballast resistor - they lower the dwell time at low RPM to lower the heat load on the coil (which runs on full voltage).  The chip in the control module just keeps raising the dwell time as the rev's rise - Jeff Morgan (my favorite local mechanic) told me he has seen some of the HEI chips do 35° at higher revs - and as low as 12°-15° at idle. 

I wouldn't be surprised for Accel to have engineered their points replacement unit off the HEI chip (or similar). What sort of total coil / ballast resistance was recommended with your unit? Maybe the dwell time it is allowing (at idle) is too small compared to the resistance in the coil primary circuit (on the edge of too weak a spark) - but then as the rev's (and the programed dwell) go up, things begin to work better.  

You might try disconnecting the vacuum line to the distributor and plugging it temporarily on the carb side.  If the idle smoothes out, you have too much ignition timing at idle which could simply be a result of using a direct vacuum source for the distributor versus ported vacuum.

If you’ve ruled out the ignition timing, I’ll suggest you revisit the idle mixture screws at the carb.  Experimenting with the idle mixture when the engine is hot is easy enough to try and gut feel says the engine is lean so opening up the idle mixture screws some more would be a starting point.  Are both screws turned out equally or are they staggered.  From the closed position, how many turns are each opened.

On the four barrel carbs, the rear float level is important as there is a fixed idle circuit on the secondary side of the carb that also adds fuel to the overall mixture.  Shooting some carb spray into the eight air bleeds located on the fronts and rears of the venturies is also worth a try.

I’ll also add that too wide a spark plug gap can also make for an intermittent miss at idle.  You can try 0.025” gaps at the plugs and see if that helps in the short term.

Frank 

You could make a gauge to check your valve lifts, since a few thousands difference lobe to lobe wouldn't be that big a deal in this case.  A strip of metal, plastic, wood, with a notch to sit on the valve retainer and the end cut to length to touch the head surface would work.  Turn the engine until one valve is fully open, adjust the length of your "gauge", and check the rest of the valves at the fully open position.  Or use a dial caliper as a "depth gauge".

http://books.google.com/books?id=G2kUibvRUo8C&pg=PA8&lpg=PA8&dq=valve+lift+dial+gauge&source=bl&ots=mjmwsTFVIq&sig=

K2YIZogD7vDUEEWn55W1si3xrDw&hl=en&ei=rwmvSqXbOoOHtge7762bCA&sa=X&oi=book_result&ct=result&resnum=8#v=

onepage&q=valve%20lift%20dial%20gauge&f=false

There are probably more than one valve fully open at at time, but I haven't researched it.  I would probably use a remote starter button, hold the gauge against each spring down the line and bump it over until the "leg" touches the head, then go to the next one.

Though there is some variation in lifts (did you notice if exhausts typically differed from intakes, could be the cam was made with more lift on one) I don't feel it is enough to cause your rough idle.  You have already diagnosed the problem, low compression on some cylinders.  Looks like a valve job is in order.  To convert cam lift to valve lift, multiply cam lift by your rocker arm ratio.  If you still have the original rockers, that would be 1.43.