Pretty certain that paintballs don't much care how or by what path they are loaded into a breech. Assuming the same force from the loader, right, center, left, vertical, diagonal, etc. makes little difference. Force is force, and the number of turns, bends, etc. prior to breech should have minimal impact on performance.

Concerning the offset feedneck and paintball bounce theory as indicated in the PE video, that is positively dubious. In fact, the video itself completely refutes that notion. After ball #1 fires and the bolt retracts, ball #2 falls into the breech under its own weight, bounces, falls again under its own weight, bounces again, and then balls 3 & 4 (damaged in the loader) are pushed down on top of it. Notice that they come down together, unlike ball #2. You can clearly see (though the paint) that after the damaged balls are shot (both #3 and #4 are eliminated in a single cycle), all the subsequent balls in the stack fall directly under pressure, and bouncing is insignificant. So #5 (covered in paint) drops with #6 directly on top of it, and almost zero bounce/dribble. #5 is fired, clearing things up a bit, then # 6 drops with #7 directly on top of it, and again, essentially no bounce/dribble.

This of course fits with the principle of force vs. gravity feeding. Having continuous even pressure on a stack of soft spheroids will pretty well eliminate the possibility of breech-dribbling, regardless of how the feedneck is oriented to the breech.

Honestly, after watching that video closely, I'm not entirely certain whether and to what extent it is actually instructive.

To the original issue, so long as you run a parabolic feed plug to eliminate undue stresses on the balls, there should be no meaningful difference in fed cyclic rate from this to a vert feed, or side feed, or whatever.

For me the only disadvantages a powerfeed has are (a) that the feed tube is open to the elements, to dirt, and most especially, to incoming fire, and (b) the need for an elbow if not running a warp/Q setup.