To be fair, my original idea was for a layer of Delrin or similar material bonded to my redesigned steel charging handle to minimize the friction resulting from the now tighter tolerances of that design. But if the material (Delrin) is capable of handling the job there is no reason not to make the entire CH out of it. The paintball experience suggests it may be as the recoil of those devices actually can be more punishing to equipment than a real firearm.
Regarding @Kona’s point, the only thing holding the CH in place and the only thing that can hold it more snugly against the bolt tube is the recoil spring and the associated notch in the CH, the design of which differs in the recoilless CH from other versions of the CH. See photo below. That fact alone could explain the difference seen in @Kona’s photo - and why a better comparison would be between two CHs of the same type - but his point remains a valid one. If the notch tolerances vary, so will the snugness with which the CH is held against the bolt tube by the recoil spring. There is little that can be done about the OEM CH tolerances, but those of the three MCarbo CH variations are completely within its control. If it maintains consistent (and tight) tolerances, each and every CH, revised or otherwise, should ride along the bolt tube at the same distance from the tube without regard to the firearm on which it is installed.
That leaves as variables only the dimensions of the bolt and bolt tube. While some variation in the bolt tube is possible, any substantial variation of the diameter (which is what mainly matters here) is mathematically improbable; at least to the extent it would significantly impact the snugness we are seeking to achieve. As a practical matter, if KT allowed substantial variation in this parameter it could not sell a replacement heavy bolt with any assurance it would fit all of the firearms sold. And if the diameter does not vary, neither does the circumference.
I do not have enough knowledge to comment specifically on the bolt itself, but relevant deviation would result only to the extent the recoil spring channel was drilled off center or oversized. We do not know the acceptable KT tolerances for that operation but I would expect them to be fairly strict.
Despite the foregoing I am concerned about @Kona’s comment that it works because the tolerances aren’t tight and will stop working if you make them tight. Do you stand by that in light of the above Logan and would you please explain if you do?