I am not sure dedicated full time employees can do better job as
whole, on a single person level a professional developer, encyclopedist or writer will outperform even the best amateurs sure, but as a community many open source projects ( documentation or applications) have outperformed professional ones many times.
PostgreSQL is a good example, it fantastic documentation and runs on mostly volunteer work the documentation is on par (IMHO better) to say Oracle/ MSSQL/ DB2 documentation which have paid professional writers behind it for example
Quality is more a function of editorial oversight rather than nature of contributions itself.
SO and wikipedia has both been effective partly because of stricter review process ( reason their brands are trusted) and also have been criticized for it, getting the balance right is not easy.
Sub-reddits are good way to study this, there are ones which emphasize on participation over quality and vice versa, i.e. optimize for the reader or the writer.
There are no perfect solution, and communities must keep changing, early on contributions are more important i.e. quantity, as the site and brand grows large quality is more important (SO and wiki both went through this lifecycle).
MDN has very different starting point - strong brand and content quality but no contributing community. They will have to come with something that fits their unique needs.