One other (often glossed over) consideration is that the redundant systems are usually considered to be independent as far as failures go but may not be.
Using orders of magnitude worse failure rates to keep the number scomprehensible a simple example is:
If each system has a 1 in 100 chance of failure (per hours/flight/whatever) in theory there is only a 1 in a million chance of losing all 3.
The 'gotcha' in this is that the failures causes are not truly independent for several reasons:
1: Already alluded to is higher load on remaining 2 systems.
2: Maintenance errors affecting all 3 systems. (Famouse case is missing O rings on oil filters on all 3 engines)
3: Manufacturing or (undiscovered) design defects affecting all systems.
Lastly is a design or maintenance error causing independent systems to not be as independent as they should be.
Others have mentioned very significant safety and operational considerations when deciding to press on or divert.
Anybody else ever cringe when some supposed expert describes an aviation event as a 1 in one million happening.
Considering total number of commercial flights per year I would hope for much better than that.