Language Changes

I have a question about language change. Is it prossible that grammatical forms
may change such that forms which have a similar phonetic realization are
replaced with forms that are more disctinct? (This may seem a bit strange at
first, but I'm going to back it up with examples)

Say for example, there is a language with singular forms of nouns like this:

NOMINATIVE            GENITIVE/DATIVE
q'dol    /c@`dol/      q'doll /c@`doL/
dhjed'   /`DjEd@/      dhidax /Di`daX/
d'rma    /d@r`ma/      d'rmax /d@rmaX/
rrach    /r~atS/       rr'chax /r~@`tSaX/
mje     /mjE/          mjex /mjEX/
ngor    /Nor/          ngorsh /NorS/

The plural forms are thus:

NOM P.                    GEN P.
qell'    /`cEL@/          qell'q /cEL@c/
dhjen'   /`DjEn@/         dhjen'q /DjEn@c/
drem'    /`drEm@/         draq /drac/
rrech'   /`r~EtS@/        rrachq /r~atSc/
menj'    /`mEnj@/         man'q /man@c/
nigr'    /`nigr@/         nugr'q /nugr@c/

Somehow, I would like to derive a system in which in a later stage of the language,
there is an opposition between singular NOM and GEN and plural COMM. That is,
somehow I want to get rid of much of the case system. What I thought to do was
take forms from either the Genitive or the Nominative plural for the new plural
form based on whether or not that form is most different from the singular
Nominative.

so:

q'dol, q'doll  >  qell'
dhjed', dhidax > dhjen'q
d'rma, d'rmax > draq
rrach, rr'chax > rrech'
mje, mjex > man'q
ngor, ngorsh > nigr'

I hope I've made myself clear :)

Elliott