0 is just as incorrect as 12157665459056929000. By the way, $calc(2^1050) already implements your feature.
Computer calculations always have a certain limitation in range and precision, if you really need better than what mIRc gives, google for big numbers or arbitrary precision math or something.
Maybe you can even write a mIRC dll using
http://spanky.triumf.ca/www/fractint/arb-prec.html or something?
