Slim Slag
There are no 'definitive' facts about engine reliability as applying to an individual, just a set of stats that allow a post hoc analysis and some regression analysis. You will find that probability will work over a big sample, but to the individual the severity of impact will be as critical and the decision will revolve around one's view on that.
Bottom line, risk/reward decisions are made on perception of the upside/downside equations, e.g. look at the FTSE 100 everyday. As the disclaimers say, past performance is no guarantee of future performance.
That's as good as it gets and trying to fit the stats to this discussions is the equivalent of using Procrustes' bed in Greek antiquarian literature - you will get the answer you seek, but it doesn't mean that it's valid.
I think you're barking up the wrong tree here - there is not a definitive answer to your question, too may qualitative variables are involved.
If you think that piston engines are reliable and that singles are safer for twins, then for you they are.