>>The question remains why do we need "fancy" expressions at all?
>
> Because linguistics can't possibly afford to have a clear and unified
> terminology
Whatever the value of that argument in general, it's certainly not applicable to this case: as far as I know, ‘lemma’ is the universally-used term for canonical (in lexicography), uninflected forms (in morphology and computational linguistics). You don't get much more unified than that.
As for clarity, I've never seen what's wrong with having a term of art for a concept that doesn't have an unambiguous correlate in everyday usage. ‘Word’ wouldn't work in this case, since the point of ‘lemma’ is grouping various inflected word-forms under a common heading. In fact, employing the everyday term can be a source of considerable confusion between the lay and the specialised meaning (as Arnold Zwicky has pointed out regarding ‘grammar’ [http://arnoldzwicky.wordpress.com/2012/02/22/its-all-grammar/] and Geoffrey Pullum regarding ‘passive’ [http://languagelog.ldc.upenn.edu/nll/?p=2922]).
A.