Biblia Hebraica Stuttgartensia notes (in its characteristically cursory way) that some manuscripts ("mlt mss") have "וְצֹפַר". Exactly which manuscripts have this form is unclear; Aleppo codex, Leningrad B19A, Sassoon 1053, Venice Makra'ot Gedolot lack this vav. I found some printed editions online with the vav, but they are of limited critical value since their source manuscripts are unknown. It would be worthwhile to consult the newer Biblia Hebraica Quinta edition, but I do not have access to a copy now.
Many important witnesses have "and" before both Bildad and Ṣofar. For example:
The Targum has:
ואזלו אליפז דמן תימן ובלדד דמן שוח וצפר דמן נעמה
The Syriac (Peshitta) has:
ܘܐܙܠܘ/ܘܐܙܠ ܐܠܝܦܙ ܬܝܡܢܝܐ ܘܒܠܕܕ ܫܘܚܝܐ ܘܨܘܦܪ ܢܥܡܬܝܐ܂
The Seputagint (Rahlfs/Hanhart ed.) has:
ἐπορεύθη δὲ Ελιφας ὁ Θαιμανίτης καὶ Βαλδαδ ὁ Σαυχίτης καὶ Σωφαρ ὁ Μιναῖος
The Vulgate has:
bierunt ergo Eliphaz Themanites et Baldad Suites et Sophar Naamathites
BT Bava Batra 15b (Vilna, Venice, Pizzaro, MS Paris 1337, MS München) and Avoda Zara 3a (Vilna, Venice, Pizzaro, MS Paris 1337, MS JTS 15, MS München, Cambridge MS. +CUL T-S AS 74.183 [Cairo Genizah]) use this form. In fact, I cannot find any editions without the "and". Incidentally, they also consistently write Ṣofar as maleh: וצופר.
Though the best Masoretic mss leave out the "and" before Ṣofar, the best quality ancient translations do have it. As such, the vav is potentially original. Admittedly this suggestion is not entirely certain and therefore is not a complete answer to your question. However, it gives a theory about why this "unnatural" and apparently unique form you note exists here.