Well? I tested my formula with 10,000 rows and it took 0.3 seconds to calculate. So, AFAIK, no performance issues. HOWEVER, I am a bit baffled by your expected results.

Please explain HOW you derived the expected results for February.