The general pattern seems to be that, if you're interested in building a better X (browser, compiler, operating system), then a monoculture is bad. If you're interested in building on top of X (websites, code, applications), then a monoculture is great, so long as the dominant entity is good enough.
In lots of cases, I think the people building on the platform tend to get their way, both because they're more numerous and because the technology in X eventually stabilises, so fewer people want 'a better X'. If a truly better X does later emerge, it then needs a Herculean effort to break the monoculture (e.g. Firefox in the bad old days, clang, the Linux desktop). Then there's an interesting transition period before, perhaps, a new dominant force emerges.
Maybe Mozilla can maintain a mixture of rendering engines by being the determined underdog. It would probably make it easier for a new rendering engine to emerge, but the open web may seem less inviting than more uniform technologies. For all its faults, one of the reasons Flash did so well for so long was that developers didn't have to deal with multiple competing implementations.