I really get frustrated when people say that one is "always" better than another. The truth is, you use it to tune your curve like any other thing on the engine.
For instance, I needed a buttload of initial advance on my inefficient 454 with 8.5:1 compression. In fact, I locked the advance at 36*. Using manifold vacuum would have put me at 36 degrees initial and retard it all the way up to WOT. That is exactly the opposite of what it needs to do. Instead I set it up with ported vacuum, so that it was 36* base, then the vacuum advance added more during low-load cruise. Respectfully disagreeing with techron, ported vacuum was NOT a stop-gap for emissions. It was a drivability choice enacted by the manufacturers given the lower compression of the early smog-era vehicles.
On the other hand, if you have a high compression engine, you can use manifold vacuum to get extra initial timing that backs off a bit under load, then the mechanical puts it back in as you increase RPMs.
The bottom line is that the vacuum signal you use is simply another way of tailoring the ignition curve to match the engine's needs. One isn't "right" and the other "wrong." This debate is much like the vacuum secondary carb vs. the mechanical secondary carb. There is no "right" answer, only what is best for your application.
With your setup (good match of cam, heads, and compression by the way) either one could be tuned for proper operation, but I would lean toward the manifold source. Set it up for 8* initial (off the can) and then most vac cans give an additional 12* or so under vacuum. That will give you 20* at idle, but still allow easy cranking. Curve to give yourself 32-34* total. Play around with total, but the Vortec heads are very efficient and don't need the 36-38* like old school heads do.
In your two scenarios (and I'll add a third), let's look at how that would play out. Both scenarios start with 20* at idle
1) 10 initial + 10 from manifold vacuum. When you hit the accelerator, vacuum drops to nearly zero, meaning you instantly drop 6-9 degrees from your timing. Depending on your stall, you might need to curve the dizzy to start adding timing earlier to compensate.
2) 20 initial and no vac advance. You would be missing out on throttle response and acceleration by not having that extra, throttle-based timing adjustment.
3) 20 initial and ported vacuum. 20 at idle, then when you crack the throttle a little, the vac can responds with additional lead which tapers off as you add more throttle.
All three scenarios start at 20 degrees. The first example reduces lead when you add throttle. The second example doesn't do anything until you reach the speed when the mechanical kicks in. The third example adds lead when you add throttle. Then, once you pass about 1/4 throttle, all vacuum sources will have the same general reading. So the fine-tuning we're describing here only has to do with cruise, part throttle, and throttle response. Idle advance is very inconsequential. Engines will operate on a huge range of advances at idle. At WOT, all vacuum signals are basically zero, so the only part of the tune you have to really worry about are light throttle inputs. The manifold source scenario will make an aggressive idle advance that backs off for part throttle. The ported source makes the same idle advance, but attacks with a more aggressive off-idle lead.
I suggest you get (if you don't already have one) an adjustable vac canister. It can be set up for both sources. Try it on both. If you hook it up to ported (starts at 20 and increases) and find that its pinging under load/lugging, try backing off the vac advance. If you find that its still pinging, try manifold vacuum and keep going with the adjustment.
While on ported vacuum, the adjustment limits how much it adds when you hit the gas. On manifold vacuum, you're adjusting how much it takes off when you hit the gas.