tdracer many of the arguments I have listened to about ETOPS could equally be applied to a case for saying that 1 engine is as safe, or safer than 2. And there are many cases where additional engines don't help, eg running out of fuel, to name the most obvious.
But you contradict yourself by asserting that "after you have two engines, adding more doesn't help safety," and then saying "there is a small but finite probability that a twin engine aircraft will suffer a dual engine shutdown or loss of thrust that turns catastrophic - and the probability gets smaller if you add more engines". Exactly.
It is a fallacy that you increase the chance of a catastrophic failure by having more engines. Apart from related failures (ie fuel starvation, volcanic ash etc) each failure is a random event. The probability of the first failure is the same whether your are flying with 2, 4 or 8 engines, as is the probability of a second unrelated failure which is, as you say, small but finite. I do believe that the small probability of a genuinely random second failure is increased by the fact that one engine has failed, due to a number of factors that are hard to quantify. For example; if the failure is the result of a maintenance error because the mechanic was tired, cross and under pressure, did that mechanic also work on the other engine on the same shift, and sign off both tasks?
The limiting factor on allowable diversion times was, if I recall correctly, the longest time that a hold fire could be suppressed, nothing to do with engine failure probabilities, and maybe that's still the case. I often wondered if that figure has ever been really tested in all conceivable types of hold fire. Maybe it has, maybe it hasn't. That's a problem with quads also, of course!
By the way; "First off, that failure wasn't supposed to happen". There you go, Murphy strikes again. It wasn't supposed to happen, but it did because it could.