Pretty much a sidebar, but during the brief time this forum got me curious about side by side coach guns last year, I learned that it's pretty simple to disable the safe-after-closing mechanism on most of them if you want to.
As discussed here and elsewhere on the forum over the years, this is the page I am on for any SA firearm with a manual safety. SA firearms without a manual safety are excluded from the field about which I make decisions, other than to exclude them. The same practice with decocking applies to TDA, of course. I'm only bothering to re-address these points because I was surprised that they did not seem to be settled in the first part of the thread.
Another point that's been rehashed elsewhere is that the literature that's available reporting on these studies does not discuss, as far as anyone here has been able to state previously, whether the actual travel/movement of the trigger was measured, or just the force applied to a non-moving trigger. An obvious human factors question that does not seem to be addressed in the literature is whether the movement of a DA trigger contributes to the resistance to pressing it all the way to the break. Conventional wisdom among DA supporters is that it does, but I have not been able to dig up any quantitative measurement reported in publicly-available literature that would tend to either support or falsify such a hypothesis. It does seem logical that a factor correlating with mechanical work would be relevant. An SA trigger that breaks at, say, even 8 lb., with just the "creep" movement, is obviously not the same thing as a DA trigger that starts at maybe 4 lb, swings through an arc of maybe 0.2-0.3 inches length at the center of the trigger shoe, and rises to a force of 8 lb before the break. The amount of energy (in the mechanical/physics sense) that must be applied to fire the DA trigger is much greater. The discussion of the flinch force measurements is typically carried out as if all that matters is the force at the break.
Training the finger to index is a requirement no matter the trigger system. There is no handling a firearm safely without it, period. The fact that the safety or decocker can be selected through choice of firearm makes the question of training for them different.
Personally, I find BBI's "single point failure" argument compelling, and have for some time.
Also, I have put in most of my work with DA or DA/SA triggers. At one point, I got hold of an SA version of a TDA pistol I had quite a bit of experience with. Same frame, ammo, mags, grips, trigger. Everything the same, except a 1911-like (and nice and big) thumb safety. I was thinking it would be my double stack 1911 substitute. Spent some time with it in dry fire and took it to the range with a shot timer. Found I was several tenths slower to the shot from low ready than I was with the same pistol in DA. Running a camera at public ranges around here isn't really an option, so I didn't have that resource. I couldn't really identify any "pause" or delay in the presentation caused by "stopping" to disengage the safety. My conclusion was that it seemed the extra cognitive layer required to deactivate the safety took away from the mental resources to conduct the rest of the action, which slowed down the entire process by the amount measured.
I considered whether I could eventually train past this. I reasoned that even if I did, doing the same amount of work to improve my performance with the system where I started out several tenths faster would likely lead me to still be at least a couple of tenths faster than I would end up being with the slower system. (My athletic abilities have never included quickness, so I'm struggling with speed in any case, and likely always will.) Although there is risk of failing to decock a DA/SA trigger after shooting, there is likely to be more time available to identify and correct that error than there will be when trying to get from the draw to a shot. With the work I've done on DA/SA transition, I often find that of the two shots in a pair, the DA is the more accurate. I'm also comfortable with DAO triggers. Especially with a bobbed hammer, a DA/SA gun is not likely to get accidentally cocked while in the holster, but there are many anecdotes of SA pistols being manipulated off safe after being holstered. Combining all that with the single-point failure reasoning, I sold all my safety-equipped SA pistols that are not Buck Marks.