My push test could have definitely been off. It seems that a lot of newbies have a hard time getting it accurate and I'm sure I'm no different.
I've done quite a bit of push testing. I've had bullets stick or drag when I remove the cartridge, producing a longer OAL than the process should have. I've had bullets push in a bit as they drag on something at the last moment and produce a shorter OAL than the process should have. There are methods to compensate for both, and per the instructions here at the site, you're supposed to perform the process a bunch until you see the same OAL, or a cluster of OALs within a range of a couple thousandths, pop up over and over again, then settle on that.
The process does work, but from time to time I'd get results that were considerably off, and at some point I decided that if ultimately I wouldn't know for sure how accurate my results were until I built a cartridge and gave it the ol' plunk and spin, I decided to come up with my own process to get to that point a little quicker.
My process is to push test exactly one bullet in exactly one case. That gives me a ball park. I then seat a dummy round -- size, bell, seat, crimp -- and adjust to an OAL maybe .02-.03 longer than the push test result, something that I know for sure is too long. I plunk that dummy and confirm, then put it back in the seating die and seat it a little shorter, a little shorter, a couple thousandths at a time, plunking and spinning after each reseat to check it. When you do this, it won't seat all the way, won't seat all the way, then it seats all the way, but still won't spin, won't spin, then starts to spin with a little drag between the bullet and rifling, then a little less drag, then almost no felt drag, then no drag. I measure and record that final OAL as my true max OAL without rifling engagement.
As long as I am only reducing OAL .001 to .002 at a time, I know my recorded true max is no worse than .002 from actual factual true max, which in all likelihood changes a thousandth from one bullet to the next anyway. I then deduct a cushion from that true max for a working max OAL.
And that's where my OAL starts for load development.
For someone using the push test and finding it to work as intended, I see no reason to change it. But if you have problems producing a consistent result, or if you get a consistent result that is out of the norm, you can skip the middleman and go to the dummy route I take.
I did have 4 FTF out of 60 rnds ( 12 mags loaded with 5 rnd each). 2 were the first bullet in the mag when I hit the slide release and 2 were on the 2nd round. The FTFs measured 1.047" to 1.053"
Yeah... one thing you can do is to stop treating the slide lock as a slide release.
I'm kidding. There are a thousand pointless discussions online as to whether or not that component should be called a slide lock or a slide release. Regardless of what you call it, I'd recommend you stop using it to release the slide. You will likely have better luck sling-shotting the slide. Load a fresh mag, pull it back all the way with your hand, then let it go to chamber the round.
Also, if your gun is brand new, you may just need more time for it to break in. Sometimes when guns are new, there are contact points between different parts of the gun that will smooth each other out over time, but when the gun is new, there's enough friction to slow the slide down just enough to cause feeding problems. You may find that sling-shotting for a while avoids the problem, and that after a few hundred more rounds, the problem no longer exists as metals polish out each other's rough spots.
Another possible problem with new guns is that there's gunk in it slowing down the slide, and a detailed cleaning will fix the problem.
But don't settle on "it's too short" because it can just as easily be too long. One of my pistols had feeding problems with the Blue Bullets 125gr RN at a particular OAL, but if I loaded it .01 shorter OR longer, it fed fine.
Food for thought.