I am no expert in CONN, but I had the same issue just now. In my case, I just converted the two conditions into two sessions for each subject, and then the Dynamic FC could be computed correctly. I think it should also be perfectly acceptable for your case since rest1 and rest2 are really just two sessions for the same subjects, so you don't need to put them as different conditions.

However, I have another study where there are really two different conditions (controls vs patients), and in this case, if as you say CONN cannot yet find the connectivity for independent conditions, then CONN cannot yet be used for this kind of study. If someone has a solution to this issue, that would be great...