Prenasalization, not ejectives cause of Winter's law?

Rick Derksen
(Old Icelandic jarpi "hazel grouse", rjúpa "ptarmigan"
and their Germanic and Balto-Slavic cognates, in
Language Contact; Substratum, Superstratum, Adstratum
in Germanic Languages)
quotes Henning Andersen for the following summary

Proto-Slavic
xxxxxxxxxx simple complex
full grade e:rb erimb
zero grade i:rb rimb

Proto-Baltic
xxxxxxxxxx simple complex
full grade e:rb erumb
zero grade i:rb rumb

of his proposal to account for substrate NWEuropean substrate
words meaning "hazel grouse", "ptarmigan", "rowan tree" and
"speckled".
Derksen emendates it to

Proto-Slavic
xxxxxxxxxx simple complex
full grade erHb e:re/imb
zero grade irHb re/imb

Proto-Baltic
xxxxxxxxxx simple complex
full grade erHb e:ru:b, erumb
zero grade irHb rub

and remarks that it is tempting to ascribe the laryngeal to
Winter's law, but that the complex Baltic form (Latv. rubenis
"black grouse") precludes that. Apart from that, if this is
a freak form, Derksen's summary shows an alternation between
a voiced unaspirated produced lengthening and a nasal, in other
words *-V:rb-/*-VrVmb-. Now suppose that voiced unaspirated are
not preglottalized as claimed but instead prenasalised, as
Pulleyblank has proposed, that alternation would be instead
*-VrHb-/*-VrMb-. That means that positing voiced unaspirated
as prenasalized can account for Winter's law as well as
positing them as preglottalized.


Torsten