The answer is of course, "it depends on your combination." Lets review the fundamentals. Timing advance is to account for the time it takes the fuel-air charge to burn and create maximum cylinder pressure. We want that to occur between TDC and about 15-20 degrees after TDC where the rod has maximum leverage on the crank. The strength of the mixture, compression, and RPM affect the amount of advance needed for this to occur. Richer mixtures burn faster, so more advance is needed for lean mixtures like idle and cruise. The amount of advance also rises with RPM up to a point, but that diminishes as RPM rises, typically staying constant in the 30-36 degree range once the engine is within the cams efficiency range.
On a street engine, we need to keep a couple other things in mind. It has to start easy, idle smooth and get reasonable fuel economy. Also, emissions are a thing. To start the engine, we want the spark to occur close to TDC. For idle, we want it more advanced, but not necessarily the optimum advance. For power, obviously we want the spark to put our peak pressure to occur 15-20 ATDC, and this is generally in the constant range. Going down the road, we want a nice lean cruise, so we need more advance, but by RPM were in the constant advance zone. So we want to advance even more but based on engine load instead of RPM. Hence the vacuum advance which is load dependent.
So what does all this theoretical BS have to do with UTG? I think he's trying to make a distributor with a totally wrong advance curve a little less wrong without spending any money or actually doing it right. Why would a distributor have a totally wrong advance curve? Back to the engine combination. For a 7.8 compression 2 Bbl 318 in a family truckster, it's not so wrong. But there we're concerned with easy starting, smooth idle, and emissions. In this case, we don't want optimum advance at idle , strange as it might seem. Why? Because carburetors. So why not advance it to the optimum point? Two problems. If your cam has much overlap (duration and/or close LCA) this will be in the 20 degrees BTDC range. Problem 1, once you get past 10 or so, the engine gets hard to start. Problem 2, carbs are airflow dependent. So the idea is to retard the timing at idle. By doing this the efficency is reduced and RPM is lower. To compensate, we open the throttle blades more. The carb likes this. Opening the throttle blades means more air moving through the carb, and it's therefore able to more accurately meter fuel, especially lean mixtures. So for emissions, which means lean mixtures, we want less advance and more throttle blade opening. Again, for a stocker that rarely goes past 3000 RPM this retarded lazy timing results in a smooth easy starting engine that burns pretty clean.
With a performance engine with higher compression and a higher operating RPM, we really don't care so much about idle or low RPM smoothness (we don't race at 1500 RPM). In this case we want a quick advance curve but need less total advance. UTG is taking the quick and dirty approach by chucking the stiff spring. So this gets the advance in quicker which is a good thing. And, you can obviously find the best point for full advance by testing with a stopwatch on a back road. But, given the fact that low performance distributors have a large amount of centrifugal advance built in, the optimum timing for power might actually be way too retarded for idle. Open the throttle more? Yes, that will help. But now starting and idle aren't the problem. Turning the engine off is. Wide primary throttle openings at idle will cause run-on issues. The crutch here is to connect the vacuum advance to manifold vacuum and not ported. So doing this will add about 15 degrees once the engine starts. I suspect this is what UTG is doing when he says 20 BTDC at idle. That's most likely with full vacuum advance at idle. 5 degrees static makes an easy start, 20 degrees will smooth out a high overlap cam, and getting up to 34 or so total makes power. What's not to like? Heck, we've all done it, and it usually works OK (unless it doesn't). And why wouldn't it? Try this with a factory low-stall torque converter and the problem becomes apparent. With the vacuum advance active at idle we can close the throttle blade opening back down quite a bit, meaning the engine will shut off when we turn the key, which is a good thing. But then when we put it in gear, the tight converter will drop the RPM, which will reduce the vacuum, which will reduce the RPM more, the carb will have a hard time metering fuel, and the engine will want to shut off when we don't want it to. Running a really light advance spring, which allows for some mechanical advance in the idle RPM range makes this worse.
Long story short (too late on that) UTG is describing what we broke kids did in the 80's.
