It is clear that the Lyngdorf output is a BIT on the high side to perfectly match the 1-1.5V input sensitivity of many poweramps. But that is not to say that Lyngdorf is at fault. But nevertheless it will result in a fair bit of digital attenuation in the MP-50 and a slightly higher noise floor than strictly optimal. It is a shame that not more power-amps provide different gain settings but several do and that is a good thing. All of mine do so I am in luck ;-) Also many high-end DACs can be configured to output similar or even a lot higher output voltages, including my dCS, and I use the 6V balanced output setting, thus requiring very little/zero gain in my preamp/input buffer for enough power for "loud" speaker SPL. If you check that previous article I referred to (
Review and Measurements of Purifi 1ET400A Amplifier), you will notice that they discuss gain matching a fair bit and also measure the impact of that on SNR. I will not claim to understand all the possible tradeoffs and implications, other than stating that it is important to consider gain structure in a chain of digital and analog components, and I always had good results with having high output voltages just after the DAC... just like the MP-50 does... as long as the amp can handle it;-)
However, you should also consider that RoomPerfect (or any RC) will also attenuate the internal signal in the digital domain, to provide sufficient headroom for correction filters (at least -6dB, perhaps more). So I doubt the MP-50 in practice will ever output 5V rms, which I assume is for a full-scale 0dBFS signal, which the DAC will never see with RP activated, even at full volume. So as a compromise, I would *guess* that amps with an input sensitivity of perhaps 2V for full power should be a good match with the MP-50.
I have not experimented a lot with this, other than reducing the input sensitivity of my active speakers to match "reasonably well" with the MP-50 output, and then I left it at that; I also think you must have a VERY quiet room for this whole discussion to become more than a theoretical issue except for cases of fairly extreme gain mismatch. I still think this is something Lyngdorf should give some recommendations on...