But are there really a significant number of actual defense shootings where the bullet was recovered and checked for expansion?
It's been in service with NYPD and, I believe, LAPD and Chicago PD. It's proven because it's done a good job of quickly stopping a threat. Massad Ayoob tracks performance based on reports he receives from police departments and recommends the SB-GDHP. Dr. Roberts recommends this load, too, and he uses a combination of gel tests and after-action reports from shootings. They have access to a large sample size and, to me, that gives their recommendations a high degree of validity. At the same time, I don't put a whole lot of emphasis on highly quantitative statistics (i.e., "one-shot stop percentages") because there're too many variables that can't be accounted for; which is why I rely on their more qualitative recommendations. That's what I base my choice on.
Whether it expands or not is not as important to me as street results, and that it can penetrate adequately while minimizing the risk of overpenetration. I have read about it expanding in actual shootings, but I can't say how much or how consistently that is.
Also, people often misinterpret gel results. Gel is homogeneous, people are not. Lots of variables affect the expansion and penetration characteristics of a round. Gel is used because testing protocols can be standardized and allow for comparison between rounds. The FBI standard of 12"-18" in gel means that, according to the FBI, a round that will penetrate within that range in gel should penetrate deeply enough to reach an attacker's vital organs while minimizing the risk of overpenetration. Many modern JHP have been reported as having textbook expansion when recovered from actual shootings, and many modern JHP have failed to expand in similar conditions. As I said before, what matters to me is how well the bullet performs in terms of quickly stopping a violent attacker, while minimizing overpenetration risk, not how consistently, or inconsistently, a round expands. To me, expansion is a benefit if it happens, but should not be counted on or expected.
Another use of gel results is to take advantage of it's consistency/standardization to compare the performance of proven rounds with unproven rounds. Rounds that perform similarly in gel
should perform similarly in actual shootings. In theory, at least. Which is why I still prefer to choose rounds with a proven track record in actual shootings.
Having said all that, other factors are more important than terminal performance, such as shootability, reliability, etc. Being able to get good, quick hits on target is more important than how well a round performs, whether in gel or on the street. But if one can shoot a street-proven, reliable round with sufficient skill to get good hits on target, why would one choose anything else?
This is my line of thinking when it comes to choosing a self defense load. I'm not saying I'm right and anyone who disagrees with me is wrong. We each have to make our own choices, and what criteria we choose to use or ignore to make our choices.
I hope that helps explain where I'm coming from.
Also, apologies for the novel.