The die contacts the "TOP" of the shell holder and the bottom of the shell holder has no effect on the amount of bump. It can effect the amount of cam over force but it will not push the case further into the die.
The purpose of the Redding competition shell holders is to change and reduce the amount of shoulder bump without touching the die.
If a standard shell holder will not give you enough shoulder bump then just lap the top of the shell holder on some fine wet and dry sandpaper a little at a time until you have the desired shoulder bump.
You can also try pausing at the top of the ram stroke for a few seconds, this will reduce the amount of brass spring back after sizing.
This (bold, above) is somewhat incorrect - true only if one uses a single shell-holder: the OP was questioning Redding Competition shell-holders producing undesirable results. In the case of these shell-holders, any variation in the height of the shell-holders can and does produce negative results. In order to establish consistent die to shell-holder contact, the distance between the top face of the ram, and the top of the shell-holder dictates that the thickness of the shell-holders, above the ram, must be uniform - the industry standard is .250" +/- a thou or so - see pic. The Redding, and most other shell-holders hold the tolerances for this attribute very close.
In my post, above, look at the lower pic: that is the web (varying THAT width is similar to the feeler gauge trick) - for that holder, .1200", or, 0.005" less shoulder set-back than standard/maximum for a SAAMI minimum spec. chamber, which would measure .125". Once the ram and die are in contact, the shell holders should be interchangeable - and, if there is NO GAP between the ram face, and shell-holder rim face, they are fully interchangeable.

The thickness of the "web" is what determines the amount of shoulder set-back: the Redding COMPETITION shell-holders work almost perfectly -
IF the
ram shell-holder slot is
NOT so narrow as to hold/create a gap between the
proper contact surfaces. There MUST be four surfaces, in contact: shell-holder bottom face to ram face; bottom/mouth of die to top face of shell-holder. Any other contact points are undesirable.
If there is a GAP between the bottom of the shell-holder rim and the ram face, caused by the bottom of the shell-holder contacting the bottom of the holder slot, any variation in the holder height and the top face of the holder, will result in that much variation in sizing.

The very bottom of the shell-holder contacting the bottom of the shell-holder slot, in the ram, is not uncommon, and is the primary reason that changing shell-holders results in vexing results.


Assuring that there is NO gap is sound practice.
If, with a static die adjustment, you cannot change shell-holders and experience/feel the same "crush", something is, "out of whack", but likely easily correctable!

RG
P.S. following this post, I measured 25 shell-holders, of various makes & case-head diameters: the OVER-ALL height ranged from .473" to .501! The rim faces, to top faces ranged from .247" to .252" - that was somewhat surprising - I expected .250 +/- maybe .0015, or, a range of about 0.003". All of my Redding Comp. sets measure between .250" and .251". The web dimensions, supposedly in .002" increments/steps, beginning at .125", are mostly close, but sometimes, 0.001" off.