Wikipedia:Redundancy is good
When it comes to the behind-the-scenes administration of Wikipedia, redundancy is good. Just as RAID arrays and similar redundant systems exist in the computer field to allow for the possibility of a hardware failure, so should the same principle apply to Wikipedia when it comes to editors. In the Wikipedia system, having multiple editors that perform the same task is useful because it allows for the possibility that one or more editors leave the project, temporarily or otherwise.
Why do we have 1000+ administrators and 30+ bureaucrats? Just in case some of them are temporarily indisposed. Redundancy allows systems to operate more efficiently by allowing for unforeseen circumstances. Computer hardware is not immune to failure, and neither is Wikipedia.
Logically, we can condense this to the following principle: redundancy should be sought in any situation where there are few negatives to having it. Systems whose value becomes that of the weakest usable component would not qualify as such as situation, just as some computer parallel memory arrays are limited by the speeds of lower-powered chips, or when many chefs attempt to prepare the proverbial broth. Wikipedia editors generally work as independent actors or on an ad hoc group basis, so in general, redundancy is good.
This may discount the reasoning "we have enough X (admins, bureaucrats, mediators...); more X are not needed." It does not necessarily discount the reasoning "we have plenty of X, so we can afford to be selective", when combined with a reason not to promote.
Redundancy engenders aggregation, and aggregation is one of the main principles on which open source and other collaborative projects such as Wikipedia are based. Aggregation permits to build on the efforts of others by tapping into the wisdom of the community rather than the wisdom of the few.