Crimp is definitely not needed on bolt rifles, but I do polish my expander down to .0025-3" below caliber size for a little more interference. You'll get better results by shooting samples of 10, that's a good idea.I think I'm going to ditch the lee crimp die, load up 10 more and roll crimp them, and see if that makes a difference. Speer recommends not using a lee factory crimp die, which I just learned today.
When FL resizing you definitely don't want to set it up to get the "cam-over" feel of the die bottoming out on the shellholder on a belted magnum. You're now resizing the headspace off the case head instead of the belt. You'll need a headspace gauge (https://www.amazon.com/Hornady-HK66-Lock-N-Load-Headspace-Gauge/dp/B000PD5VN8) or comparator kit for your calipers, and measure the case from head to datum on the shoulder, bumping .002-3" below fired dimension. If the cases still chamber easily I would continue to neck size until they're fully formed and you get resistance on bolt close, then FL resize from there. I remove the expander ball on initial die setup to not overwork the neck, and then replace it once you're achieving the correct bump. Contrary to popular belief I've never had an expander ball pull the shoulder back out and change the dimension, and I don't believe they induce enough runout to actually change the precision of a load.
Yes, typically trashed cases will be noticeable. Split necks can be visibly seen. I use a hand primer seater so you can easily tell the difference in resistance on seating. If you set your FL die up as I explained above, you shouldn't have issues with case head separation (where the bottom .25" or so of the case actually breaks away and separates from the body and neck portion), if you set it up to cam over like the instructional videos you will from overworking brass.