Somethin like a 9mm or 45 ACP wouldn't and shouldn't be roll crimped, no matter how well trimmed
And those competing with 9mm and .38 Spl in the pistol class of PPC using roll crimps for accuracy are doing it wrong?
If you're certain your pistol brass is definitively headspacing by the mouth of the case contacting the front of the chamber, go do a casting of that chamber to assure yourself it's actually touching.
And I'll stick with roll crimping my 220 grain 10mm Bear Wrench loads - whether or not the edge of the crimp actually touches the front of the barrel chamber.
Having designed and built a few wildcats, can assure you it very much does. I dealt with this same problem in a super sized .350 Legend, a .577-500 Jeffery.
Once you have no shoulder and no rim, you have one option left for headspacing; the case mouth. This is how rimless straight wall automatic pistol cartridges headspace. And you can’t put a strong roll crimp them either for reasons that will become obvious if you try.
This isn’t much of a problem until the recoil builds, as in a light carbine and rifle cartridges. Then the recoil without a strong crimp can cause bullet setback or dislodge them. It can be made to work, like the .30 carbine which the military stake crimped to prevent the issue. But it is a highly compromised way of achieving head spacing.
Unless you’re stuck to bizarre rules stateside, there aren’t many good reasons to go this way. As I discovered myself in the .577-500.
Again, PPC and Bullseye shooters using .38 Spl and 9mm autopistols have been using roll crimps over the top edge of wadcutters from long before I bought my very well used and very accurate S&W Model 52 in the mid 1970's. They're specifically doing it to get the best accuracy, meaning they're more concerned with the best and most consistent headspacing from shot to shot, versus whether or not seated bullets move during recoil.
But then, these aren't .577-500 rifles or similar wildcates; these are conventional cartridges that have been used in competition longer than the vast majority here have been alive.
If you believe you can get a consistent OAL case length for headspacing using taper crimping, but you CAN'T do the same if you roll crimp instead, I'm curious to know why. Presumably, whether you roll or taper crimp, the end result is the OAL of the case after crimping is the same within a thou if you properly trimmed your cases beforehand.
Then you can explain to me why the roll crimp I put into hard cast heavy loads for my 10mm Dan Wesson CBOB aren't actually working despite the fact I've owned it for almost 20 years now.
I didn't invent the idea of using a roll crimp instead of a taper crimp for either application with autopistols, BTW - it far precedes me coming along and following what they learned.
For fun and games, drop some empties that were taper crimped and then roll crimped (minus the bullet being seated) into the chamber of your auto pistols and then use a bore scope to see how many of them are actually properly headspacing: i.e. the edge of the case is tight against the front edge of the chamber.
My Dan Wesson came with match chambers in both the .40 S&W and 10mm barrels and neither touch the front edge of the chamber. Ditto the Model 52, although it's the closest of the three barrels. My Browning High Powers aren't even close in comparison, ditto my wife's S&W Shield.