Rubbing away the shine (1)

Does anyone else find the PIE reconstruction of *mer- 'to shimmer, to shine' suspicious? For one thing, there are many meanings being given to this single root, each with its own series of cognates. Porkorny listed five separate meanings. The blogauthor of Bradshaw of the Future noted that the American Heritage Dictionary combined 'to die' and 'to rub away' into a single etymon, something that makes perfect sense to me because it would suggest that 'to wear away' is the original sense from which 'to die' is to be derived through metaphor.^[1] Presumably these two meanings would have been distinguished by two different conjugations: objective *mḗr-ti 'he rubs something' versus subjective *mór-or 'he wears away; he dies'.

Back to the meaning of 'to shine', it seems out of place. Nonetheless, Mallory and Adams in the Encyclopaedia of Indo-European Culture (1997), page 514, tries to assure the reader that the 'shine' root is "widespread in IE". Yet a closer inspection of the evidence shows something suspect:

1. Not one cognate shows the unreduplicated verb root and meaning directly.
   (ie. All are obscure derivatives.)
2. Many things are left unexplained.
   (eg. How is Sanskrit marīci- 'shining mote' to be analysed morphologically? Doesn't the unusual pattern of reduplication in Greek μαρμαίρω [marmáirō] deserve explanation?)
3. Taking each branch separately, the verb is weakly evidenced.
4. Some of the cognates can equally be interpreted using a different sense of *mer-.
5. Added all up, there are an unusual number of odd problems demanding special pleading.

Is this really a genuine PIE root or is this falsely reconstructed from language scraps? Furthermore, is there a foreign non-PIE root lurking here that might be gleaned from this mess? More later.


NOTES
^[1] Puhvel, Hittite Etymological Dictionary, vol.6 - Words beginning with M (1984), p.150 (see link).