Do you mean I can't state that I find A's soundstage is deeper than B without offering measurements or blind tests?
I did not limit you to those options either. The standard is what gets presented at an engineering/scientific conference such as AES, ASA, IEEE Spectrum, etc. Other types of explanation accepted there are references by other researchers, measurements, psychoacoustics, etc. Just a flat observation with no way to analyze it is not a topic there so won't be a topic here.
If I say that I believe this may be due to small signal non-linearity do I have to now produce measurements? Can I not state that this is my speculative hypothesis ?
I don't know how the word "believe" goes with something being "speculative hypothesis." Putting that aside, I am afraid you still can't say that. Simple reason being that the concept of "small signal non-linearity" and "deep soundstage" is not an accepted concept in audio engineering/science circles. You can talk about small signal non-linearity by itself and that would be fine. But connecting it to an audible observation is not the direction we want to go here. I hope you understand.
Maybe it would be useful if you listed the "accepted standards of audio science" ?
You had to put me on the spot, didn't you?
Accepted type of communication are:
1. References to research backing it. Like to see references and quotations.
Example: "side reflections can improve listener enjoyment. See Toole, Olive et. al.
2. Measurements.
Example: "DAC A has lower distortion, noise, jitter than DAC B as shown in this graph." Graph can be your own measurements or third-party.
3. Listening test results.
Example: We did a blind listening test where we matched the levels and my friend and I took turns and both of us could tell 8 out of 10 times which unit was which." Alternatively you can link to references by others.
4. Design and architecture.
Example: "PCM is better than DSD because it has less in-band noise due to required noise shaping in DSD."
5. Engineering excellence.
Example: "Use of an isolated USB interface reduced the opportunity for interference by the source."
Again, I want to be clear that these are the requirements and examples when trying to advocate a point of view. Nothing wrong with saying you swapped your speaker cables and think that the soundstage seems deeper to you. We get in trouble if you say "speaker cables with silver interconnects have better soundstage than copper."