is there "i" suffix that makes verb from noun? etymology of tio(n) suffix

Question

Wiktionary says on PIE -h₃onh₂-:

Descendants

Italic:
...
Latin: -iō (from *-i-h₃onh₂-) (e.g. legiō (“group of selected people”))
Latin: -ō (e.g. Nāsō (“having a conspicuous nose”), poss. Iūnō (“having heavenly authority”))
...
Latin: -tiō (from *-ti-h₃onh₂-)

The -tio suffix:

dictātiō (“a dictating, dictation”), from dictātum, supine of dictō (“I dictate”)
quadripartītiō (“a division into four parts”), from quadripartītum, supine of quadripartiō (“I divide in four parts”)

You see, this article does not say that dictātiō is from dictō, but it says it is from dictātum, supine of dictō.

Supine:

There are two supines, I (first) and II (second). They are originally the accusative and ablative forms of a verbal noun in the fourth declension, respectively.

Fourth declension:

Nominative portus
...
Accusative portum
...
Ablative portū

Inflection of dicto:

non-finite forms - participles - passive - perfect : dictātus
verbal nouns - supine
- accusative : dictātum
- ablative : dictātū

They have common part dictāt, (but most of them has common part dictātu), I can conclude that maybe supine of dictō has stem dictāt, and dictātiō is dictāt + iō.

Most forms of "dicto" have common part "dicta", so I can conclude that maybe dictātiō is dictā + t + i + ō , where t is passive perfect participle suffix and ō makes agent noun from verb according to https://en.wiktionary.org/wiki/-o#Latin . so, "i" should be a suffix that makes a verb from a noun.

Was there really "i" suffix that makes verb from a noun?

Or maybe really dictātiō is not from dictātum, but directly from dictō (infinitive dictāre) as dictā + ti + ō, in that case what is meaning of "ti" suffix that does not consist of t + i?

Answer 1

The usual explanation goes back to Benveniste 1948:

(Benveniste 1948: 97).

It is still the most common analysis that can still be found virtually everywhere, e.g. Fruyt 2011 writes the following:

"-tiōn- fem. is a Latin adaptation of inherited *-ti- (following the zero grade of the root in other IE languages, i.e.the short form of the root with no media e), phonetically reinforced by *-ōn- or *-iōn- in order to avoid phonetic disturbances" (p. 159).

In other words, in many cases Lat. -tio is a secondary, athematic suffix.

However, Pultrova 2011 questions the analysis mentioned above.

She dismisses Benveniste 1948 proposal (that PIE *ti > Lat. ti-o) on the grounds that:

"the subst. in -tiō probably have nothing in common with the inherited subst. in *-ti-, as is generally believed.”

She argues that “Lat. subst. in -tiō originated through secondary derivation by the stressed suffix *-i̯én- from the verbal adjectives in *-tós“ (p. 85).

She also mentions a very interesting proposal by Olsen and Rasmussen 1999, who claimed that the suffixes -to-, -tu- and -ti- are allomorphs of the same morpheme.