Mersenne prime discovery rate
 

The expected ratio of successive exponents is 1.48.
Each discovery will need 1.48^3 = 3.25 more computing than the previous one - more iterations, more time per iteration and more candidates to test.

Let us say such an increase in LL-computing occurs every four years.
We can expect a prime every four years for as long as this "Moore's Law" continues.

The happy reason I am pointing this out now is that this is now close to our current expectation: 0.1% chance of a discovery in well under two days.
Note the modest resources needed to accomplish this:
Ten LL tests per hour, taking a month on average to complete.
30*24*10 = 7200 tests in progress at any one time.

David
[SPOILER]One of the 7200 non time-wasters:smile:[/SPOILER]