Two things:
a) how many actual failures of engines in certified airplanes to you really hear of - as opposed to the discussions of this possibility?
b) for various reasons, everything in light a/c has evolved incredibly over the years - just look at the avionics available (although not necessarily installed). The one thing that still lives in the 1950s is the engine - with the concomitant reliability. The interesting thing is that nobody has managed to come up with (and certify!) anything better - just look at Thielert.
PS: by 'failure' I mean a mechanical malfunction - not pilot-induced a la running out of fuel, not applying carb heat, etc