Duly respecting members better endowed than this humble soul
but what are we designing here?
an instrumentation amplifier??
I find a specification of within 0,2dB rather thick for something audio. It seems to be accepted that 1 dB is already pushing it as far as audible differences go. And then in direct comparison; not a signal today and the same 1dB better tomorrow. Then we come to valve production spreads: +/- 10% is - er - acceptable; see posts elsewhere of what the real picture out there is. Rough approximation: This will account for a spread in rp and thus gain of some 1 dB in this case.
Phono cartridge response vs. frequency? I can assure you that within +/- 1dB is pretty darn good; thats another possible 2dB of variation.
I found the circuit of post #18 giving the RIAA graph to within about 1dB with a bogey valve. Again, as always, you spend your money as you please. For myself I cannot see myself paying boutique money for 1% capacitors and such.
Input capacitance worries you? Use it as part of the RIAA h.f. cut. In fact, I have seen RIAA circuits for MM cartridges using an input capacitance only as the h.f. compensation, reportedly also giving superior h.f. performance of the cartridge. (Though I will place that 2nd best; one then forfeits the advantage of also attenuating fist stage hiss in the process.)