I'm at best just a seat 'o the pants hack with the electronicals, learn as I go just enough to get the job done (or not).
It would be feasible to use most any old externally triggered coil and rig a pickup of some sort (optical, magnetic, hall) off the flywheel. Coil could be powered off the flywheel but chances are it won't easily adapt so you'd need a battery to power an external coil.
Then use any of a number of adjustable/programmable control modules available/intended for car/m'cycle use to adjust the timing on the fly. Or physically move the pickup, lots of options...
Integrating it into a dyno chart would certainly further streamline the optimization part. That would be Eric's
@malk315 domain, but he's got more than enough on his plate with dyno stuff for the time being...
Problem with not having to adapt a coil and/or pickup to every different saw and using the stock self contained/triggered coils is that the SCR or whatever that triggers the field collapse event is potted in and you'd need access to that stuff to alter/interupt/replace it, typically not fun to dissect without mucking things up.
Best shot at real time adjusability short of replacing the coil with an externally triggered one would be a mechanical advance, like on those old outboards, physically moving the coil.
Saws don't usually have a lot of extra real estate to do that, some do though.
No simple answers...
On a side note, I've found in my travails of adapting somewhat more modern ignition systems (ike GM HEI) to older m'cycles that some types of very simple magnetic based pickup coils seem to provide an inherent advance curve as a byproduct more or less of their physics. As rpm increases, the triggering threshold advances along with it, some sort of 'effect' evidently.
I've also found that there is generally only a very small benefit to complex advance curves on basic engines as opposed to an almost immediate ramp up to full advance at relatively low (read starting) rpm.