Not sure what you are trying to do here. Are you arguing that ὁ ὢν in Romans 9:5 is in the third attributive position ? Because that is impossible.I ran this through a search (since there is no source) and it appears your quote is not from a grammar, but a blog, "Growing in Greek," and that is located at http://growingingreek.blogspot.com/2011/07/substantival-participle.html.
The Blogger writes (as you quote), "A substantival participle is an independent use of the adjectival participle. It is used instead of and functions in the place of a substantive." You are accusing me of ignorance in the language but seem to be fundamentally misunderstanding something you apparently read on a blog. In what way specifically does the nominative singular masculine ὁ ὢν not accord in person, number, and case with the nominative singular masculine ὁ μονογενὴς υἱός? And how does it function independently when it stands immediately next to the noun it modifies? Are Harris and Metzger also surprisingly ignorant?
As Wallace notes, I quote, "The adjectival participle may occupy any of the three attributive positions and both predicate positions. You should normally translate the attributive participle as though it were a relative clause." (Greek Grammar Beyond the Basics, p. 617). That's because the function of the clause is adjectival. Note:
"Relative clauses are also sometimes referred to as adjective clauses, because they identify or give us additional information about the subject of the independent clause they relate to." (Grammarly)
Since you won't take my word for it, you will note that Wallace in his grammar specifically offers ὁ ὢν in John 1:18 (note the variant) as an example of an adjective in the third attributive position, and after noting this example he adds:
"More frequent than the adj. in the third attributive positions is the participle. When a participle is used, the article should normally be translated like a relative pronoun." (Greek Grammar Beyond the Basics, p. 307)
Also, if Wallace really argues that ὁ ὢν in John 1:18 ( with the variant reading μονογενὴς θεὸς) is in the third attributive position, such is a highly suspect take. The way most ( if not all) Trinitarians argue the grammar with this variant is by taking ὁ ὢν εἰς τὸν κόλπον τοῦ Πατρὸς as an appositive . So this is hardly a clear, and irrefutable example from the GNT of ὁ ὢν in the attributive position ( first, second or third). Not to mention that the θεὸς reading is already highly suspect.
So you just have no precedent in the GNT for what you are trying to do at Romans 9:5 with ὁ ὢν.