One quick clarification: FAA's policy is: if there is technology available to meet the reliability numbers in AC 23.1309, the reliability numbers are a requirement; if there is NO technology to meet those numbers, then the maximum that can be achieved is acceptable. Obviously piston engine technology does not allow an engine which, as a unit, meets the 10^5 criteria. Therefore the attainable piston engine reliability today - 1 failure in 10000-13000 hours - is acceptable. But the idea is to do whatever possible to meet those numbers. That AC is common sense. As for turboprops and turbofans, that is a whole different ball game. P&W PT6 reliability average is around 1 IFSD in 330000 hours, and big turbofans such as the GE90 (Boeing 777) are 1 IFSD in more than half a million hours. Interesting to remember that, 65 years ago, turbo compound radial engines (such as in the DC-7 and in the Super Constellation) could barely meet 1 failure in 1000 hours!