Tom,
I'm no "tuner", but I've messed with my Edelbrock unit a bunch and I can tell you what worked for me. From your comments I'm assuming that your Holley ECU is NOT self learning, and that it is programmed via 3D maps and correction curves. I can tell you I was lost until I bought a wide band O2 sensor that I could mount in the exhaust with no leaks, feeding a digital fuel/air ratio indicator I could watch for tuning. Mine was more difficult because of the short lakes style headers and no tail pipes. With a passenger watching the numbers and taking notes, you drive around in your "problem areas" of the map. A data logger would make it quicker, more accurate, and much easier, but an indicator works. Now this assumes that you have your Spark Map nailed down, and that you're not making changes on two fronts at one time - that'll get you mired in the mud quicker than anything if you're chasing a problem.
The key to getting mine "happy" was the built in correction curves that modify the map based on coolant temperature, and it sounds like that may be your problem, too. I got my Fuel Map where it was working right for WOT, then cruise, then worked the transitions. The F/A vs coolant correction is basically the "choke", but the curve goes all the way out to extreme hot temps, too. This is your cold idle, transition through mid-range as it warms toward normal, and then compensates above normal just in case. The one that really kicked my tail was crank fuel vs coolant temp, which is the amount of fuel injected to get responsive starting in all temperatures - mine was veeeeery sensitive in the transition area where the curve is near vertical.
It sounds to me like your guy has the spark and injection maps pretty well set, but that he may not be spending enough time on the correction curves, assuming the Holley uses a similar interface to that of the Edelbrock unit.