[morzh]

The document is a bit confusing, when coupled with actual datasheet:

https://suddendocs.samtec.com/productspecs/ipx1.pdf

But OK, they tested it to three times the 750V for DWV. Datasheet uses confusing terminology, for WV it says "DWV" but these are different things. If 750V were DWV, the WV would be 250V. 

Anywho. Just remember, again: this test is for connector itself. Not mounted on the PCB. A clean plastic with controlled humidity of the air will make it to pass pretty high voltages.
However your worry now is your PCB. This connector, soldered to a PCB will not withstand long-term 750V, let alone over 3kV.
So. You probably do not need to meet 750VDC as WV, as maybe 500V is short-term (a few seconds) event, that exists while the tubes are warming up, for a few seconds. Then it will go down to maybe 450V or lower for the first rectifier cap.
But you can meet 500V if you 1) use SMT version, not through-hole (and make sure the solder pads do not protrude too much inside the footprint.) 2) pot the solder side with corona dope or what they have today for electric breakdown varnishes.

Better yet, use IPBS/IPBT connectors if you have space (you should). I think they should meet the same voltages as the otehr one, but over 4mm pitch will make it much better for surface breakdown. And, again, it never hurts too pot the solder side.
And, BTW, with these you could use through-hole versions as you will still have 100 mils board distance after using 65 mil pads, which is much better.