I have to say, I am curious why you, and to be honest many other on this forum, but of course you can only answer for yourself, are so quick to assume it is a support issue whenever there is a dimensional stability problem?
Back on the Form 2, I was working on some directly printed injection molds, and as a result did extensive testing on dimensional stability and warping. On that printer, I found no strong correlation between number of supports and dimensional accuracy of my test models. The number of supports obviously had a strong correlation with things like delamination and complete part failure, but it was my experience from a large number of calibrated test prints, that once there were sufficient supports to successfully print the model, there was literally no increased accuracy to adding more supports, unless you were specifically talking about a very tall part, that extended quite a way up from the platform in Z… Orientation, the type of resin used, the condition of the optical window, tank, and mirrors, even position on the platform, all could have measurable changes to dimensional accuracy. But if the part would successfully print with say 50 supports, doubling that to 100 supports, would not make the part any straighter, it would just make the support side uglier,
Now, there might be something I am missing, and I haven’t done this same sort of specific testing on the Form 3, However, after years of printing I have found that if a model is warped, or out of spec, it is rarely, if ever, going to be fixed by just cranking up the number of supports. Sure, if it is missing from the supports, or part of it didn’t form, or part of it delaminated, obviously you need more supports. But if it is misshapen, warped, or out of designed spec, there is usually some other problem.