I'm not an expert on this, so this may just be a case of the blind leading the blind, but here's my take. I'm going to do a little bit of thinking out loud here, I apologize in advance for the rambling. Maybe someone who is an expert will chime in and throw some cold water on this. But, nothing ventured, nothing gained, so here goes:
If you are wiring the H295 motor as bipolar parallel, then based on the typical formula for inductors in parallel, the net inductance of wiring two 3.3 mH coils in parallel would be about 1.6 mH.
IIRC, the formula for two inductors in parallel that I used assumes that there is no mutual inductance between the two inductors. I'm not sure that that condition holds true for the two stepper phases that are being wired together.
If indeed there is mutual inductance between the two phases, then the formula would need to be modified to take that into account. Maybe there's a reference somewhere that addresses the mutual inductance issue within steppers, but I haven't run across it yet. I would note that Jones (Jones on Stepping Motor Types) says that 8-wire steppers are wound with bifilar windings - if that is truly the case, then I'd expect that indeed mutual inductance would be a factor, and so my simple inductors-in-parallel calculation is probably wrong, because it does not incorporate the mutual inductance factor.
If I'm reading it correctly, Jones talks about how to determine the operating voltage for bifilar windings in series, but not in parallel - unless you consider the following quote:
It is worth noting that some permanent magnet stepping motors have 4 independent windings, organized as two sets of two. Within each set, if the two windings are wired in series, the result can be used as a high voltage bipolar motor. If they are wired in parallel, the result can be used as a low voltage bipolar motor.
So, one way to interpret that would be to simply take my original approach of treating the two windings as independent inductors in parallel, giving a combined 1.6 mH inductance, which is what I will do here.
Also, I do not know the assumptions behind the commonly-stated suggestion for choosing a voltage that is 32 x the square root of the coil inductance in mH. So, maybe that would be a good one to research as well, to see whether those assumptions hold true in the case of wiring the phases for bipolar parallel.
But if you take that as a given, then the square root of 1.6 is something like 1.26, and so 32 * 1.26 would give a result of about 40 volts for the recommended operating voltage.