Gotcha. Seems like everything about this disease is "it depends."
More like "incomplete information" than depends, IMO.
There are numerous studies in various aspects of celiac disease...but the more you read the more you realize most of the studies have skewed participant bases (because they're based on Celiacs who were primarily dxed based presenting with outdated criteria like "must have D", etc.". Then, you look at sample size, and dropout rate, then the term "statistically significant" creeps in.
All of these criteria are typical for medical studies; however, if there's one thing AI diseases are NOT - it's typical.
So, what you're left with are mostly studies on Celiacs who presented with "typical" celiac disease, tested positive on blood and biopsy (which isn't always achievable), and who are probably not "super sensitive" and can tolerate a degree of tinkering by medical professionals: being able to successfully change diet, successfully tolerate a gluten challenge, etc.
And in children....wow, the skew just gets more extreme since its harder to get an early + on blood work and that trickles down to fewer biopsies...which trickles down to fewer dx.
Then introduce NCGI....which helps but muddies at the same time. There are probably quite a few Celiacs carrying the NCGI label because one or more tests were -. Sometimes testing was done well, other times it was done poorly or not at all.
Gluten sensitivity dx and understanding is abysmally incomplete. More is known each day....but it's slow.
IMO what is REALLY needed is a test that can be performed WITHOUT a gluten challenge. Since medical advances and understanding of gluten sensitivity are on fast-forward these days, and the only way to manage gluten sensitivity is a gluten-free diet...retesting or investigating where an individual falls (Celiac or NCGI) requires a cruel and destructive event - a gluten trial. Not many are willing to do this. So, the already dxed (with a wealth of knowledge and successes and failures of the diet, lifestyle and challenges) are reluctant to undergo repeat testing.
Most of the best study participants don't wanna do it. And I am one of them, and I don't think any of us are making the wrong decision.