Yockey’s estinmate for cytochrome C is a good start.
You’re underestimating by more than 75 orders of magnitude.
the second problem is with convergent evolution. if we will use the number from talkorigin (10^40) for a specific function then it means that to get the same function again by convergent evolution is extremely unlikely. so we need to expain it too.
That number isn’t from talk.origin, it’s yours. So the only necessary explanation is that your estimates of probability aren’t reliable.