So says Charles Lee Irons, “Let’s Go Back to ‘Only Begotten,’” Gospel Coalition website, 23 Nov 2016: https://www.thegospelcoalition.org/article/lets-go-back-to-only-begotten#_ftn3

Irons begins by noting that in the KJV there are five Johannine passages that speak of the “only begotten” Son of God (John 1.14, 18; 3.16, 18; 1 John 4.9). He then notes that in the modern era there has been a broad scholarly consensus that μονογενής means ‘one of a kind.’ He then accurately represents the rationale for this consensus: “Scholars have argued that the compound Greek adjective is not derived from monos (‘only’) + gennao (‘beget’) but from monos (‘only’) + genos (‘kind’). Thus, they argue, the term shouldn’t be translated ‘only begotten’ but ‘only one of his kind’ or ‘unique.’”

Irons offers as his first argument that μονογενής means ‘only begotten’ in some passages. This presumably means that there is no noun like ‘son’ or ‘daughter’ in the context to already suggest birth, though he does not say this. It is certainly what I expected in order for his argument to make much sense, however. Otherwise, ‘one and only son/daughter’ makes perfectly good sense, which would defeat his point.

Irons begins by citing one reference from Plato—Critias 113d: μονογενῆ θυγατέρα ἐγεννησάσθην. Here not only is ‘daughter’ mentioned explicitly, but also that she had been ‘born.’ If μονογενής here means ‘only begotten’ then an awkward tautology occurs: “They begot an only-begotten daughter.” (The Attic aorist middle dual is here used.)

Further, I was surprised to read his three biblical examples:
Luke 7.12: μονογενὴς υἱός—here ‘son’ is explicit.

Luke 8.42: θυγάτηρ μονογενής—again, explicit.

Luke 9.38: διδάσκαλε, δέομαί σου ἐπιβλέψαι ἐπὶ τὸν υἱόν μου, ὅτι μονογενής μοί ἐστιν. But here ‘son’ is already mentioned, so the ‘one and only’ [son] is simply good economical Greek style.

Thus, Irons’s approach so far is simply question begging.

He follows this up with 1 Clement 25.2 [Irons says it is 25.1], which speaks of the Phoenix as ‘one of a kind’ using μονογενής. He also mentions an unidentified text (‘an ancient treatise’) that speaks of trees as ‘in one kind.’ But he adds, “these are uniformly metaphorical extensions of the basic meaning…” That, too, is begging the question, because he is assuming that the essential idea of μονογενής has to do with birth.

Second, he says that “careful examination of the word list of Thesaurus Linguae Graecae reveals at least 145 other words based on the –genēs stem.” This is a more significant argument, but I would need to see his evidence before recognizing its validity. He also adds that “fewer than a dozen have meanings involving the notion of genus or kind.” To argue from other words that have the –γενής stem as though they must inform the meaning of μονογενής may seem to be imbibing etymological fallacy, especially since there are some –γενής words that have the force of ‘kind’ or ‘genus.’ However, if ‘begotten’ is the routine meaning diachronically, and especially synchronically during the Koine period, Irons may well have a point.

He does seem to engage in etymologizing, however, when he says that γενός and γεννάω “both genos and gennao derive from a common Indo-European root, ǵenh (‘beget, arise’).” He finishes his arguments by again claiming that –γενής essentially has to do with birth. The BDAG lexicon allows for the meaning ‘only begotten’ for μονογενής but seems to view this meaning as secondary. In addition, they note that in the Johannine literature “The renderings only, unique may be quite adequate for all its occurrences here.”

 All in all, Irons is right to focus on the data provided in TLG for this certainly expands our knowledge base of the term. But that he seems to have focused on cognates that have the morpheme –γενής rather than the specific usage of μονογενής, both diachronically and synchronically, is a weakness in his argument.