I think there is enough information to take back to Carberry and raise the question with them. I am an engineer (albeit a software one!), and when faced with a situation like this, we begin with a Problem Statement.
1. It seems that (at least) some Carberry plates are machined differently to the original plates, such that pressure is being applied through the cam gear shims and onto the cam gears themselves.
Also
2. It seems that (at least) some examples of the Carberry plates are machined such that they don't sit exactly square to the crankshaft.
The original design has the cam gear shims serving two purposes:
a. to prevent the cam gears grinding directly on the end plate
b. to reduce end-float of the cam gears
Therefore in removing the shims, are we not trying to cure a fault with a fault?
On point 2 above - that (at least) some examples of the Carberry plates are machined such that they don't sit square to the crankshaft), the answer appears to be shimming the plate itself between the crankcase and the three 'bosses' the the plate's retaining bolts pass through - with different amounts of shimming at each boss. This shimming could correct the plate being off axis with the crankshaft.
In fact, this shimming could address both points above.
Can we assume that the machining for the new bearing is correct, if it is off in other areas?
Perhaps we could come up with figures that show the difference in the machining between the original retaining plate and the Carberry one. If these figures are consistent across examples, then there is a 'batch' machining issue. If they vary, then there is a quality / consistency issue.
I read somewhere that earlier Carberry plates were re-machined from original plates, to take the new bearing. Are these ok? Is it just the anti-vibration plates Carberry manufactures from scratch that are problematic here?
Questions, questions.