You can, but not using the same circuit for the low voltages.
Yes, I'm thinking about having a different path for low voltage and HV. In fact, I'm thinking about developing the HV part in a modular way so that it can be added as an option. Of course, that requires the module to be prototyped and tested.
... used to individually test the resistors for TC ...
Like Conrad Hoffman said, I will keep current low at least at high voltage (0.5 mA at 1kV is 0.5 W, 1W if I go to 1 mA). Hopefully this will help mitigate TC. I read that VCR (voltage coefficient) particularly effect thick film resistors and that specialized HV film resistors use a different internal geometry to mitigate that effect. Does voltage effect wirewound resistors as well and as much ?
HV is something I don't know all that well. I have never worked with anything above 150 VDC and it scares the hell out of me. When I start prototyping with HV, I'll make sure to be very careful. I'm worried about the feedback loop, connecting 1K output to the input of a precision op amp, even with a divider looks scary to me.
Is this a one-off, or production?
Yes it is a one-off product. I don't intend to market this product. If I get a decent design, I will certainly make it open source but only if the result is good. I'm at a very early stage and perhaps I should have waited to have something to share before posting. I'm supposed to get my DAC today so I can start evaluating figures like noise and accuracy. Unfortunately, it comes with an ADR445 voltage reference which is probably not the best for my application. I'm thinking more about using an ADR1399 but that's a discussion for the Metrology forum (which have already been discussed quite a lot).