If the PIE verbal ending *(Ø)-néwti produced transitive perfective verbs.
Are all Greek -νυμι verbs transitive?
Short answer: there are verbs with middle endings that are not transitive, like πέπνυμαι “be conscious” and πτάρνυμαι “to sneeze”
Probably better to refer to this as the *-néw- athematic suffix, the -ti is just the 3rd singular ending. That wiktionary entry is a bit short-sighted, I think.
Sihler (section 455) says that this form is “copiously attested, indeed productive” in Hittite, and “forms both deverbatives (causatives) and denominatives (factitives), so hu-is-nu-zi ‘causes to live’ and e-es-har-nu-ut imperat. ‘make bloody’”
This sense is “detectable elsewhere in some formations that appear to be old, notably G ὄρνυμι ‘stir up, incite’… cf. intrans. L orior ‘(a)rise, move’.”
“In G the suffix is common, and occurs in obvious neologisms like δείκνυμι ‘point out’… There is no semantic flavor traceable to the affix [emphasis added]” He adds that οἰχνέω might be a thematic version of an original *-neumi
If *deyḱ- is the imperfective form of the verbal root, why was an affix creating an imperfective transitive added?
I would not rely on Wiktionary for your Indo-European historical linguistics. Before Hittite was learned there wasn’t really a consensus on -new-, most people thought it was just a reanalysis of the stem klneu- which is actually a nasal infix in the stem klew-. Furthermore, the word δείκνυμι, as Sihler has pointed out, is a neologism in G and so the original meaning in PIE doesn’t really apply. Whereas in an old word like ὄρνυμι, it does, and as already pointed out this has a causative meaning rather than an imperfective one.
It is probably more accurate to think of *deyḱ-/*deik̂- as a root and the affix as marking the present tense. Greek verbs in -νυ- lose this affix in other tenses. Similarly to how you have λαμβάνω in the present, ἐλαβον in the aorist. Or ἐλαύνω/ἔλασα.
If you’re interested in this kind of stuff, delve deeper into historical linguistics! you’ll learn a lot and have fun too.
Wikipedia is a convenient presentation (format) of material. I see that the following is missing:
Just looking at Hoffner and Melchert’s Reference Grammar today. In sections 10.15 and 10.16, they make mention of the section 455 that you quoted from, and give a number examples of both types.
Having no mention in the Wiktionary of forms derived from *-néwti in Hittite is a noticeable omission.
Did the nasal infix and nasal suffix originally have different functions?
The infix looks similar to the Hittite infix -ni(n)- a causative marker. Did it have that function in Proto-Hellenic or in the parent PIE too or was that a shift/development in Hittite?