Hi Dunc,
Hitachi builds (or at least used to, as I haven't seen a recent production) a very good design with very high quality. Without actually seeing it up close and personal, I'd say not to modify the armature. Unless there is a problem. Otherwise keep the comm true (tir) and the rotor well balanced. Best bearings with proper lube goes without saying. Keep it cool as possible. Hot parts are more likely to move about.
I can't guarantee it but I think 6krpm would be doable.
On Yab's armature (R.I.P.), I recall detecting some build anomaly in the photos he posted at the time. Don't remember exactly but like not seeing proper impregnation in the slots. He has a point about the radius of the parts. But I've seen many where the comm has exploded and copper still secure in the core slots. I've also seen the end turns (opposite the comm end) mushroom out. Basically, often, the area or part which fails first on overspeed is the one with the flaw. With ever increasing RPM, each/all parts will fail. More often than not, the mistake goes first. So initial design and build quality is so important.
I've seen Jim Husted turn grooves in the comm, wind carbon fiber filaments and impregnate. Can't say it saved any comms. Don't know. I certainly would not do that to a steel v-ring comm; maybe to a molded core comm.
My thoughts. BTW, I've had larger rotors to 12k, but were induction cage type with oil lubed bearings.
Cheers,
major