[bryondowd]

Just couldn't resist nerding out about probabilities. Summing failure rates would potentially get you above 100% chance of failure. What you'd really want is the product of success rates. So, if you have a 1% failure rate on a module, and you have 50 modules, that'd be 0.99^50, which brings you to about 60% success rate, or 40% failure rate overall. For 500 modules that drops to less than 1% success rate.

But yeah, agree that the odds of failure are probably lower pulling in a handful of monoliths that are well exercised than hundreds of miscellanea of varying robustness.