Re: The problem
Whilst at a purely code level there may be some merits of dropping these, from an end-user point of view it makes no sense whatsoever.
Removing NPAPI will mean no Java, Silverlight or other plugins that require this, but won't remove the most crash-prone plugin out there (Flash). It won't affect home users too badly, but it will destroy any inroads Firefox has made into enterprise. Firefox is offered as an alternative browser where I work (a very large multinational company). However no NPAPI and hence no Java means no Oracle financial system, which will make the browser worthless here.
For XUL, although the code is difficult to maintain, powerful extensions have been one of Firefox's party pieces in recent times. Dropping this will cripple the extension market for Firefox, and I cannot possibly see how removing one of your biggest features/USPs can be considered a positive thing from an end-user point of view. If Firefox only supports the same extensions as Chrome, why use Firefox?
Firefox has been losing market share for some time, and I don't think stripping out features and cloning your main competitor's interface is the solution.