My comment was about efficiency and did not use "energy return" until you did, which confused me.

I also have advanced degrees and worked in academic/federal research for 7 years as a scientist/engineer before my current role, so I'm no stranger to working with measuring equipment, data processing, and the variables between. Seems we're still able to confuse each other. :)

But your statement "literature does not say its guaranteed and the amount you'll get is highly variable" was the basis of my post saying that you need to do proper testing to really know the benefits and which shoes work for you. An outdoor test using Stryd/HR/TPE is not that, and Nike might make the slowest shoes for you specifically (in response to "just get Nike" posts).

Since you said the shoes are designed for certain types of people (50kg 2:00ish pros), I also wanted to highlight that someone like me (85kg 3:00ish amateurs) can still find 4% efficiency gains between shoes.

Somehow we have managed to confuse each other whilst seemingly agreeing without knowing it :-)
I agree with all of this although I would query whether a typical runner really has access to the equipment needed to be able to assess them to a robust enough level of precision. In a paper I have on this topic that's currently under peer review, that was something I alluded to but what was clearly stated by my reviewers.

Yep I get that - but purely out of genuine interest, what was your testing precision like when assessing the baseline shoes ? Can you share some values ?

Well, I have raced in a pair of SEP1 for a couple of years now, about 200 miles. I wasn't sure if they would be fast enough for my somewhat undertrained, certainly a couple of kg over race weight, body. I was able to snag a pair of SEP3 at somewhat of a discount - still not 'cheap'. Today I ran them head to head on the track. 1 mi warmup to the track from home in the SEP1. The plan was two laps (in lane 7) of each shoe as ABAB at 8:00/mile pace. That is about 1/2M pace for me. The test was fine with each pair of laps within a couple of seconds. It was difficult to determine if the SEP1 or SEP3 was 'better' using HR or power (Garmin). Mostly the same, but plenty of small difference that make me conclude that a treadmill might be better for this. BUT, I finished with a 1 x 400m at RPE of 9/10 for each shoe. There were clear differences in favor of the SEP3 for this n = 1. Much faster, tiny increased average HR (1 BPM) and the same power. FWIW. I'm convinced that the impact of supershoes is best assessed in an environment as close to controlled as possible...