Microsoft have released a new free ECC library which is up to 5x faster than with the traditional elliptic curve P-256 (which was so far the most commonly used curve in practical applications) and a also up to 3x faster than a well-known alternative Curve25519.
Bitcoin elliptic curve belongs to the category “special therefore suspicious” and have been so far a bizarre sectarian choice subject of serious controversy. Should this one consequently be called “double special” and therefore “double suspicious” (instead of 1 special multiple it has 2 special multiples)?
Not sure, Microsoft cryptographers do not want to admit that such curves could be very weak, and take the (same) already controversial idea of using curves with fast endomorphisms as those which can have a very efficient implementation to a new level.
After all so far there is no really serious attack known on these curves and NIST curves do not have good press either.