I have made most every type of intake over the years and tested most all of them, A lot of motors I used and built could not use true venturies so I made fuel post venturies, not quite as good as a real venturie ( not a venturie tube) but much better than a restrictor.
I would debate what "better" and "best" might mean in stunt. I might believe that the flow for a given choke area and delta-pressure could be better on what you refer to as a "true venturi" in bench testing and probably provides the most power. But I can get any amount of power I want by other means, up to and including doubling the size of the engine.
What I think is far more important is consistent fuel draw and atomization in-flight. In this I think, based on some pretty lengthy and extensive experiments, that the fuel post/spigot is far and away the best, the "through the middle" spraybar (what you would call a restrictor) is second, and the plain venturi (with a flush fuel outlet) is a (distant) third. I think the problem is the tendency of the fuel to dribble along the wall of the venturi below the boundary layer in some cases, and not in others. Injecting the fuel away from the wall, beyond the boundary layer, and in the turbulent flow downstream of the injector (either a spigot or spraybar) seems much more consistent, even if it is less "efficient" in terms of flow. David has even seem a case where the fuel was seen to form a drop right below the flush inlet and break off periodically, when the engine was running. Every time it dripped, the engine went rich for an instant. That's in a static bench test where it would seem to be perfectly ideal conditions. In the air, with the relative wind changing all over the place, and the acceleration changing all over the place it would seem to be far more likely to do something inconsistent.
In testing, that's pretty much what seems to happen. An example (of many) - I had an engine with a venturi with flush fuel inlet. In flight, it ran mostly fine, but was prone to breaking lean on outsides, and rich on insides. Tank shimming to screw up the level flight speed, but bias it for the maneuvers, sort of worked but never completely removed the issue in squares, and was so far off that it was grossly the other way in level flight and rounds. In particular, it was exceptionally prone to breaking into a hard 2 in the top outside leg of the square 8, driving the airplane towards the ground very strongly. After a tip off from David (via Billy) I made a spigot venturi of the same choke area, an replaced the original with no other changes. Now, it was about the same in level flight, much more different inside/outside rounds than it had been, and grossly richer in outside squares than it had been. I put the tank back to get the same level flight speeds, and the inside/outside difference had almost gone away completely. Once again, the same choke area, same power, same fuel/plug/pipe length/compression, timing, no huge tank shim, but dramatically different and vastly better in-flight. After doing that time after time, I am convinced.
Additionally, I also got a very dramatic improvement in consistency from merely changing the fuel outlet diameter. On the Jett venturi, I changed the "dribble hole" from maybe .100 to about .040 and it got drastically more consistent. That tells me that the flow around the hole is a very critical factor and that it matters tremendously how the fuel is injected. My theory is that the large hole provides no way for the fuel to get off the wall, since the tiny amount of fuel in the huge volume of the hole just causes it to pool at the bottom and pour over the edge. The small hole stays filled, and provides at least some momentum to get off the wall and into the stream. A spigot forces the fuel to always get injected into the main part of the stream and prevents it from getting attached to the walls, and assures that there is mixing and atomization because of the turbulence.
The same sort of effect also explains why you might want a long inlet (straightens the flow better over various inlet conditions, more consistent) or a diffuser (always turbulent flow, so the external flow field changes it less).
I also think that the spigot gives better, or even much better, fuel draw for a given amount of power. But I don't have any data other than just how it feels to go on. That's not the same as the most flow for a given choke area or pressure differential but I think it's more important. I do know I can use a MUCH larger choke area with a spigot venturi than a plain venturi with a flush outlet before it gets unstable, and it seems to have more power pretty much proportionally to the choke area. It would be relatively easy to test but I just haven't done it.
So, with respect, I am unconvinced that a plain venturi with a flush outlet is better for the things I care about.
Brett