When undertaking a comparative study of an underanalyzed language family, one should understand its systems of alternations, if any, before one tackles the greater complexities of analyzing entire linguistic structures.
(An analysis of the verb sets would alert the researcher to the certainty that many other words in Samoan have lost a final consonant.)
Consider the following forms from Spanish, spelled phonemically rather than orthographically: One pattern of inflection shows alternation between /o/ and /ue/; the other type has /o/ throughout.
Also, positing such a history violates the principle of parsimony (Occam's Razor) by unnecessarily adding a complication to the analysis whose chief result is to restate the observed data as a sort of historical fact.
That analysis gains plausibility from the observation that the neutralizing environment is unstressed, but the nuclei are different in stressed syllables.
A pre-gradation phonology can be derived for each of the three groups by internal reconstruction, but it was actually an innovation in the Finnic branch of Uralic, rather than the individual languages.
That fact would be missed if the comparanda of the Uralic family included as primary data the "degraded" states of Finnish, Estonian, and Sami.
A primary split is, in principle, recoverable by internal reconstruction whenever it results in alternations, but later changes can make the conditioning irrecoverable.
However, no denominative verbs (those formed from nouns like to gut, to braid, to hoard, to bed, to court, to head, to hand) are in Type II.
There are no verbs of Latin or French origin; all stems like depict, enact, denote, elude, preclude, convict are Type I.
That evidence shows that the absence of a "dental preterite" marker on roots ending in apical stops in Type II reflects a more original state of affairs.
Other exercises of internal reconstruction would point to the conclusion that the original affix of the dental preterites was /Vd/ (V being a vowel of uncertain phonetics).
As oddly as it might seem, the loss of the stem vowel had taken place already whenever the root ended in an apical stop before the first written evidence.
Some of them are examples of Indo-European ablaut: pendō "weigh", pondus "a weight"; dōnum "gift", datum "a given", caedō "cut" perf.
ce-cid-, dīcō "speak", participle dictus, that is, inherited from the proto-language (all unmarked vowels in these examples are short), but some, involving only short vowels, clearly arose within Latin: faciō "do", participle factus, but perficiō, perfectus "complete, accomplish"; amīcus "friend" but inimīcus "unfriendly, hostile"; legō "gather", but colligō "bind, tie together", participle collectus; emō "take; buy", but redimō "buy back", participle redemptus; locus "place" but īlicō "on the spot" (< *stloc-/*instloc-); capiō "take, seize", participle captus but percipiō "lay hold of", perceptus; arma "weapon" but inermis "unarmed"; causa "lawsuit, quarrel" but incūsō "accuse, blame"; claudō "shut", inclūdō "shut in"; caedō "fell, cut", but concīdō "cut to pieces"; and damnō "find guilty" but condemnō "sentence" (verb).
As happened here, reduction in contrast in a vowel system is very commonly associated with position in atonic (unaccented) syllables, but Latin's tonic accent of reficiō and refectus is on the same syllable as simplex faciō, factus, which is true of almost all of the examples given (cólligō, rédimō, īlicō (initial-syllable accent) are the only exceptions) and indeed for most examples of such alternations in the language.
In short, during the atonic shortening, the tonic accent was two syllables after the affected vowel and was later retracted to its current position.
However, words like division and vicious (compare vice) have lost a syllable in the first place, which would be an insuperable obstacle to a correct analysis.