Really you have to work out sensible limits and then choose the kind of methodology you are going to use...

If your mechanical shaft/disc is spinning every 5uS it has a rotational speed of 12,000,000 (twelve MILLION) rpm! This I would like to see... (I'd stand there with a stopwatch and measure the length of time until your mechanical contacts disintegrate and bearings turn to dust!). If you are measuring Wind-Speed, there is a direct correlation between the wind pushing the Cup, the Diameter of the Vane and the shaft rotation. Assuming a total vane diameter of 200mm, a single pulse per rotation every 5uS (and not taking any friction into account) a bit of math calculates a wind speed of 281,100 mph (two-hundred and eighty-one THOUSAND)... methinks a pitot tube would struggle with that one!!!!

Now, you really only have two choices - and each has it's limitations...

(a) You COUNT the number of PULSES in a given set time... this method is good for FAST Data, but falls apart for slow Data (eg you're counting one second but it takes 17 days for the shaft to rotate!).

(b) You COUNT TIME between the start of one Pulse and the start of the next... this method is good for SLOW Data, but falls apart for fast Data because of the speed limitation of the PICs.