I hope the discussion in this thread does not equate more engines with increasing safety- it isn't so, and it's been pretty well demonstrated. Besides, an operator's safety culture, maintenance program and pilot training are much more important in the safety equation. Issues in any of those areas are magnified with aircraft complexity.
I'm not saying I don't prefer multis- I've got sufficient time in them and in the GOM to know there's considerable advantages. But they come with considerable cost and twins are not appropriate in some circumstances. More complex aircraft present more complex pilot challenges, even assuming safety culture and maintenance issues are satisfactorily dealt with. Consider you're offshore and have to pull one offline... Did you consider this in your fuel planning? Or- Is having a beach landing option a dispatch requirement? If you answered "no" to either of these, and there's nowhere you can safely put the aircraft single-engine- that second engine is an illusory safety enhancement. Yes, there are circumstances where a forced landing to the water in a single is safer than dealing with OEI twin work, not to mention the other system complexities in multis, more complex systems also equal increased probability of failure.
Increasing the pilot crewing has a demonstrated safety advantage. Well?
If I was operating a multi, anywhere:
I'd have a cat-A capable twin, that I could operate cat-A as appropriate and with company support;
My multi would be as simple to operate and robust as a single without sacrificing twin redundancies (seperate fuel tanks, electrical, fail-safe power-train, etc.);
Throttles on the collective and arranged as in a 222;
I'd be dispatched only when I had some place to go OEI;
Proficient technical help on-site dedicated to the on-site aircraft only;
No time critical dispatch interfering with maximized good multi-practice;
And an autopilot or an SIC, even in a VFR assignment.
Strikes from this hurried list are arguments for single engine operation.