The disappearance of *-s -- The saga continues

As my never-restful brain contemplates daily the saga that
is Indo-European 24-7, I started thinking more on this
pesky disappearance of the nominative after phonemes *y,
*r, *n and possibly *l. As always, I came up with something.

What if *-s has an allophone [S] after these phonemes
through some process of assimilation. Thus, [S] > ZERO
(disappears) while [s] > *s (remains). Perhaps it
has something to do with alveolarization? Another idea
I had was that *r, *n and *l were dental while *s was
alveolar. So perhaps the interference between the two
areas of articulation caused the added friction of *s,
producing [S]. Then *y would also cause friction through
palatalization, likewise producing [S]. Over time, [S]
would be very prone to disappearance.

However, I'm not sure what modern phonotactics would
say about a group of dental continuants with a single
alveolar sibilant. Is that possible elsewhere?


= gLeN