Yeah, I don't see the huge improbability here.

Given that we know that:

- Fleming lived next door to an unsecure mycology lab.

- The temperature during the time period was low enough that if Fleming had left a contaminated culture unattended and non-incubated, he would have had a very high chance of getting the results he became famous for...

Well, given that the probability of discovering penicillin in those conditions is pretty high (say, if he forgot/neglected to incubate one out twenty batches, a 5% chance), and the prior probability of discovering penicillin any other way is extremely low (otherwise other scientists would have found it), bayesian calculus says the stroke of luck hypothesis is probably correct.