Hi
First: I highly appreciate you ask our opinion !
That's scientific behaviour, for sure !
I'd go this way:
1. SINAD HDMI in, speakers out @5W 4 ohm
Also same with TOS-link to speakers
(it looks like HDMI is often polluted by various "noise", and most TV set have TOS link too for sound)
This is the major criteria for AVR and should be the basis for their ranking, IMO
You may test that in stereo.
If you are able to test 5.1 or 7.2, that's a plus.
But I guess the stereo front is most important for Sound Quality in critical listening.
2. Typical DAC to pre-out, if available
As you said, mainly for diagnostic
From TOS-link and/or SPDIF and/or USB and/or Roon, depending what's available.
Not the full test for all, but mainly to check what works correctly.
This should show how much improvement to expect with an external amp. As you proposed, hanging the PureHifi may even give a figure directly comparable to 1, so that looks like a good idea too to me. Maybe more work though, and typical DAC figure is probably good enough.
3. Test analog in if available.
Mainly ADC for SINAD and frequency response.
Mainly sanity check.
3. Pre-in to speaker vs power and @5W
Or whatever path gives the best idea of power amp section performance.
Same than 1.
How good is the amp section?
Mainly useful if the figure in 1. Is so-so but 2. Is reasonable.
Clear area where cost savings may have a quite severe impact.
4. Noise level impact of a room correction, if that can be compared.
Is it possible to enter a fix correction for comparing them? As an example, by using a pre-build signal instead of a microphone input to force all room EQs to apply the same correction under 500 Hz?
Because in my measurement of miniDSP Dirac processor, the impact is clearly much higher when some correction is done. (25dB impact on SINAD in that case)
5. Headphone amp if available (nice to have)