Impedance Mismatch ???

Larryi, that might be the case but when he designs whole amp (instead of using Icepower modules) most likely he has option of setting 10k or 40k input impedance without extra stage. It might be electrical noise issue since higher impedance input is more likely to pickup interference. Preamp side is low output impedance, but not at the radio frequencies where it might be tens of kiloohms. It is tradeoff between performance of the best amp and convenience of matching for lower end amps. It is feasible that, as you suggested, he targeted class D amps for use with tube preamps while his new model 625 is class AB.

It would be a good question to ask Jeff Rowland. I'd be interested in the answer.

There are always tradeoff issues in designs. I'm sure Rowland know what his are for each particular product.

usually impedance is measured at 1000Hz signal that may show different values.

Viridian, thanks for the link - very interesting. Current loops like 4-20mA became at one point industry standard for their inherent immunity to noise or resistance of the cable especially on very long runs. Current source has very high output impedance making more difficult to induce currents in the loop.