Off the cuff, my impression of the unicontext article was that for all or most phenomena unicontext is invoked to explain, there's an existing explanation that works (or mostly works), and doesn't have the downside of needing to explain the enduring importance of non-universal contexts.
