Interesting approach, I saw a discussion recently about how mimicking real traffic tends to work better than randomizing it, since randomness becomes its own fingerprint.

Does this still blend in well with normal HTTPS traffic?