Hello,
You are exactly right. There is no way it will come out of the bracket with normal input, even at max throw. I found some carbon shards between the bearings, which is a tell-tale sign the edge of the carbon caught in that small space, flexed forward and the next input was such that it forced its way out. I am going to disassemble it on Monday and double check everything to see if could be something else.
The control output was such that the swash would have been rotated CW approx 30~45 degrees. Down elevator resulted in forward and right. When I put left in, it just went forward. The control arm on the left cyclic servo had rubbed against the aluminum servo support/ frame brace. It was not a single impact, but a smooth removal of finish from the control rod about 10~15mm in length and the corresponding spot on the aluminum was indented; which tells me I had a few seconds of the control rod moving up and down before the crash. The only way the control arm could lay against that brace would be if the lower/outer swash was rotated the said 30 to 45 degrees CW.
I guess the silver lining to this, other than the obvious that everyone around was safe, is the TT conversion is coming out and I can switch to that instead of buying all of the belt components and then the TT later. Hopefully I can get my hands on one in the next week or two, as I have a demo at the OSU Speedfest the end of April.