Yes will get that variation due to the variables that I mentioned above.
1. Variations in the amount of lube
2. Variations in the dwell time at the top of the stroke (it takes 3 or 4 seconds for the brass to quit flowing and take a set)
3. Number of strokes used (you will see variation with one stroke and almost no variation with 3 or 4 bumps at the top ) When using multiple bumps rotate the case about 120 degrees between strokes. Pull the expander through only the last time if you use an expander. It goes without saying that a springy C press or other light duty press may vary all over the place. Even so multiple passes in a C press is better than only a single pass.
4. Speed of a single sizing stroke if you only use one cycle
5. Brass temper or amount of work hardening.
6. Finish or surface texture of your die. I can polish a die a bit better than most factories because I just take the extra time and I had a lot of practice on a lathe.
7. The finish of your brass - clean brass with a tan patina has a good bit more drag than jewelery finished brass.
8. A mickey mouse case like a 5.56 should have less variation with a heavy duty press than say a badly expanded 30-06 from an old Browning machine gun.
9. Finally if you are using brass that was fired in multiple different rifles then the case expansion will vary from case to case and the resistance to sizing will vary also. This will give the same results as too little lube or sizing too fast. It just stretches the press a tiny bit until you unload the pressure on the ram.
For years I have been forming cases from a 5 gallon bucket of USGI SL-54 30-06 brass that was machine gun fired. I once put a .0005 dial test indicator on my Rockchucker and found that it makes the frame stretch .002 temporarily. A second pass (just a slight retract and bump with about 3 seconds dwell) does not show any more stretch of the press. It only takes a few seconds. To defeat the extra friction with a second or third pass.
This list covers some of the variables in your technique and your brass and dies. When each and every case gets multiple sizings you will quit seeing variation. If you only use one press stroke and have a good measuring tool you will be able to find the variation in your technique.
I get best results with a FL die that has the neck lapped out so it takes only .0015 to .000 expansion to provide a good neck grip on the bullet.
Sometimes with a large groove diameter, the right brass and the right cast bullet a trim die will work right out of the box since many are larger than a FL die in the neck. I found this out on a case forming project where I formed 100 new .303 cases to a perfect fit for a M95 Dutch Mannlicher. Part of the process development involved sacrificing 15 or so cruddy once fired cases. After every sort of crush and fold and swapping out a number of dies I could form the junk cases easily. Then I started on the new brass. I measured every case as I went. I would get one now and then that was .001 to .002 long. Maybe one out of 10. A long case was immediately recycled and 100% of the time the shoulder moved back that additional ,001 to ,002.
I use this method on my ordinary rifles both hunting and military. The bolt guns get set to .000 clearance. I can feel a slight drag closing my stripped bolt with a pinky finger. Lever, pumps, and autoloaders might need a little clearance. .002 seems to be fine for use at the range. My brass lasts a long time and my ammo seems to be more accurate. I shoot a lot of old rifles with iron sights so I am not talking about 1/4" groups but I get very good iron site groups.
My bench rest rifles have a tight neck chambers and I use an ordinary Wilson neck bushing die to size the case about .002 for those rifles.
Thanks guys. Interesting enough, I'm getting about 2 to 3 (.002 to .003) thousandths in shoulder bump variation when using a FL die on once fired brass. The 3 thousandths is about one in every 15 cases so the 1 to 2 thousandths is more common. Is this variation common?