A large number of accepted solutions in short time certainly raises suspitions, but we have to be careful.
I just checked this forum to see if anyone was annoyed by my recent activities, when I resubmitted 50+ of my C++ solutions using a different compiler (g++ 4.3.2 instead of 4.0.0-8) to check if there was any improvement in performance. This kind of stream of submissions would look like cheating to almost everyone following it.
By the way, there were some improvements, but they were not consistent. I reached the first place in 5 problems, anyway.