Ok, in my room, at a distance of 9 or 10 yards from the far wall, I braced my 507k / G48 on the bathroom door and held it on a circular target (a clock) the other end of the room. I have it set for circle only. I centered the doughnut on the clock, and then moved only my head side to side, and up and down, to the edges of the display window, trying to see a shift of the doughnut in PoA on the clock center. There was none.
I did the same test with my 507c / G34, which is set to dot only. Same results; moving my head side to side and up and down if the dot is centered on the target, I don't see the dot move anywhere, it just sits in the middle of the target, throughout the envelop of the 1X glass of the RDS, edge to edge and top to bottom. In other words the dot is on target no matter where I look.
So in conclusion, I don't observe / see "any" parallax in my two Holosun RDSs. I suspect if I mounted the gun in an vice, and used a calibrated target, and had a bunch of people take observations, over a number of samples there would be some parallax, even if a little (it seems hard for me to believe it's possible to get to 0.) But based on this, and unless I am missing something (could be) I would describe mine as essentially "parallax free" at this range. My G48 was direct milled by PM, and my 34 uses a FCD plate; maybe these are low relative to deck height over bore, which may help (I confess I dunno). But at least if parallax is there, it's well below the significance of my poor trigger control, when it comes to getting an effective first time A zone shot. Obviously this is highly unscientific, non-calibrated, just a random observation of one sample, etc.