The two papers linked in your article are such a funny contrast between transgenderism and real science.
For starters, the paper on knee surgery that was linked to in the article that claimed to know about what the regret rate for knee surgery is never actually tells us what the regret rate for knee surgery is. Because is it a paper on the dissatisfaction rate for knee surgery. Which is a different thing to regret. You can compare the two papers defintions of their key terms here:
Literally, the majority of the participants that reported dissatisfaction with knee surgery reported that the surgery was overall beneficial. Dissatisfaction in this context means that the effects of the surgery are still causing them some level of unhappiness, not necessarily that they wish that they had never had the surgery to begin with.
Compare that with the definition of regret provided by the paper on GAS.
Regret is the explicit feeling that one would be better off had they not had the surgery. In other words, the majority of people who expressed dissatisfaction with knee surgery would not meet the definition of regret provided by the GAS paper. So it's a complete apples to oranges comparison to pretend that one group of people saying they are more unhappy than they expected to be after receiving a surgery is the same as another group of people who explicitly stated that they believe their lives would be better off had they not received a given surgery.
But that's more of a failure on the reporting, where the author at Gendergp failed to understand (or pretended to not understand) that two different papers were measuring two different things rather than a flaw with the papers themselves. There is nothing inherently wrong with measuring either regret or dissatisfaction, though I do find it interesting that I have never seen an attempt at measuring dissatisfaction rates of transgender surgeries. I'm not saying such a study doesn't exist, just that I've never seen it.
Where things really fall apart is when you compare the methodologies used to arrive at these numbers. First let's look at the knee surgery paper:
The authors attempted to contact all 356 patients who had received total knee athroplasty at three hospitals in Sweden. Now, when conducting a survey, one of the most important things to look at is the loss to follow-up. You want this number to be as low as possible, as a high loss to follow-up can add bias to results, often because there is a specific group of people who have been lost, causing the remaining population to become biased.
The knee surgery paper reported a 2.3% loss to follow-up. Which is well within the 5% you hope for with surveys like this. And those who were lost to follow-up were not included in the final numbers when comparing satisfied patients with dissatisfied patients. By contrast, the GAS paper completely failed to report on what the loss to follow-up even was (though reading between the lines, my guess is that it was most of the patients being reported on), their methodology seems to have been to wait around and see if anyone contacted their doctors to express regret. Anyone who didn't explicitly do so seems to have been counted as someone who didn't regret their surgery, even if the doctors made no attempt to contact them to see how they actually felt about their surgery.
Given that the majority of people who detransition do not report their detransition to their doctors (this paper claims 76% of patients fail to do this) this is already creating a sampling bias so large as to make any conclusions the paper draws unreliable. But when you look at what they actually did decide to count as regret, it gets downright ludicrous.
The only people who were actually counted as having regretted gender affirming surgery (which in the context of this study means everything from facial cosmetic surgery to genital reassignment surgery) in this study were people who contacted their doctors in order to get their surgery reversed. This is such a ridiculously narrow definition of what "regret" is that if the same standard were applied when measuring the regret rate of knee surgery, then the paper on knee surgery would show a regret rate of 0%, as it didn't report a single person trying to get their surgery reversed.
So when advocates of transgender surgeries claim that these surgeries have a lower regret rate than other surgeries, you have to include the asterix that this is only true if the definition of what regret for transgender surgeries is is so narrow, and the methodology for measuring that regret so sloppy, that if you were to apply the same standards to the surgeries they are comparing themselves to, then those other surgeries would have a regret rate of zero.
For me, the interesting contrast between transgender surgery paper and the knee surgery paper was the difference in a research team that is actively trying to find as much information (including negative information) about the effects of their surgery as they possibly can, and a team that is trying to cover their own asses by manipulating statistics to make the surgery in question seem like it has a far higher rate of positive outcomes than it actually does.
If anything, I'm surprised they get anyone contacting them at all to try and get their surgery reversed. Given how little they seem to care about their patients, I'm not sure I would trust them to reverse the damage they did in the first place.