Oium < *aiva-?

O. N. Trubac^ëv. 'Indo-Arica' v Severnom Pric^ernomorje. Istoc^niki.
Interpretacija. Rekonstrukcija. In: Voprosy jazykoznanija, 2, 1981.
Moscow. Nauka. Pp. 3-21.


p. 12
[Even the scanty remains of Crimean Gothic in a very late 16th ct.
record by Busbecque display a single, but socially a very important
loan, the source of which is likely to be connected with Indo-Aryan
(in this case, with Tauric) and not with Iranian: Crim.-Go.
marzus "nuptiae". This word has been already compared with OInd.
márya- "young man", "bridegroom" [...] One can assume, that the
Crim.-Go. form on –z- reflects Indo-Aryan *marj- with the Prakritic
change y > j, which Indo-Aryan trait makes the idea of an Iranian
source still less plausible.]

p. 16
[Very little is known about numerals; we may hypothetize about *aiva-
"1" on the ground of the name for an island – Eon (Pliny), Oium
(Jordanes) and also AIAIH NHSOS in Homer, literally "island +
island", and the "island" is probably from original "the lonely one.]


What do you think of these etymologies? In particular, is Oium < IA
*aiva- really to be preferred over the traditional explanation of it
as dat.-loc. pl. *Aujôm?