As a crude generality a dual plane intake will "want" a larger carb than a single plane intake. Viewed as a single event (no such thing in reality), an intake runner demand will only "see" two barrels of the carb on a dual plane while it "sees" all four barrels on the connected plenum of a single plane. An open spacer connects the two sides and gives the active runner a view of both sides of the carb at the same time it adds plenum volume. It also helps even the air/fuel ratio from side to side. This is assuming that the engine wants an even air/fuel ratio and that it wants the added volume and that it wants the added carb area....you get the idea.
You can spend lots of time expounding theories on carb spacers, but it usually far better to just clip them on a try them. Things like an air/fuel ratio variance, vacuum in the intake at WOT (1.5 ain't really all that bad...put a 600 on it and it would probably be over 2.5), and a different than expected peak power or TQ RPM can clue you in that an opportunity for improvement exists. But it does not mean that anything is really there - or that chasing it is worth the cost in power at other points in the power band for a given application.
I am with you lock stock and barrel. However after bolting them on, I would say that the spacer results are saying something. Assuming there really is a problem and it isn't common with the smoothing calculation being off, the 4 hole under the carb with another spacer sorta working best lets me back into two things. Of course this is all opinion, and I can't say I ever looked at "unsmoothed" or "smoothed" data, or known which I was looking at, so maybe there is no issue at all and this is a display issue. However,
1 - The engine showed it wanted a little more plenum (not surprising at 600 hp and that RPM) Letting it pulse and breathe across both sides should dampen things a little or at least move the violence to a different RPM. Not only from a plenum volume to let it breathe deep, but also from a dampening ability
2 - Something was screwing with the boosters that was corrected with the 4 hole. It could be a top problem or a bottom problem. If it was a harmonic of valve events or a reversion issue due to some other vale timing voodoo, the combination of plenum and a longer "carb throat" due to the 4 hole could have helped the booster by smoothing the flow beneath it. Would have been interesting to try an old 2 inch 4 hole there to see. It also could have been a sloppy 850 venturi combined with it, it also could have been emulsion without quite enough air or slightly too small an HSAB or a fudged one as Blair said. If the booster is going "solid fuel" for any reason (an overstatement), it would cycle too. Which a more stable flow from the 4 hole may slightly allow the depression under the booster to stabilize a little
I don't mind the 850 design although they can be fussy, If I noticed it myself, big IF, I likely would have started with cleaning the HSAB gently, just to see if one was gunky. However, I tend to blame harmonics, as I eluded to before an 11 inch runner tunes about 7000 rpm using Ramchargers theory. This is riding the harmonics NOT airflow. I don't know the exact measurements of the port, but it's in the ballpark of being right on the edge of something at the RPM the data showed
Again, not being a dyno guy, Jay's update threw me off game a little because I am not sure I have ever known when I was looking at smoothed or unsmoothed data and how much of a smoothing factor most guys use. However, if it is an issue, my gut says maybe try a little carb checking, if not, it's a resonance issue of some sort