The problem I see with the 5033 is that the QC appears to be poor. I have some with grossly off-center filaments in the glass envelope. That can't be good for the beam quality unless you can get the filament centered in the reflector. For something like an FM Ver2 Deep Reflector, I think that will be hard to do. The reflector opening is about 8.5mm diam, while the 5033 glass is about 8mm diam. Thus, you have only 0.25mm leeway on either side to adjust the filament position. I don't think that's enough.
That said, I haven't tested the 5033 yet in any of my incan Mag mods. I do have a Mag440 that drives the bulb at 18.0V, which is also good for the 5033. The reflector is an FM Throwmaster with a 12.7mm opening. So I probably have enough space in this case to get the 5033's filament centered.
Edit: I just tried the 5033 in my Mag440 with VLOP Throwmaster. The beam quality is fair. There is noticeable "structure" in the beam. I inserted the Hikari only part-way into the Kiu socket so that the filament could sit up as high as possible (the Kiu socket is set up for an Osram 64440 IRC, which is a very tall bulb). Thus, the beam structure isn't because I ran out of Mag head travel to focus the beam. I think the deal is that the 5033 filament is quite large and is going to be hard to focus nicely. The axial filament of the 64440 IRC produces a much nicer beam with a tighter hot spot.
I also tried the 5033 with a VLOP Ver2 Deep Reflector. Same result. The hot spot has some "structure" to it. I doubt that it would be noticeable against real world backgrounds, but you do see it against blank walls. Finer filament bulbs like the WA1111, WA1331, and WA1185 have much nicer beams with clearly tighter hot spots when using the Deep Reflector.