Rather, I should say that there is no evidence I have seen that the Greek optative resulted from a similar process, though theoretically the origin of the optative could be similar. Romance language conditional moods derive from the use of the infinitive and habere as an auxiliary verb in vulgar Latin of the late Empire. The shortening of the forms is what produces the "i", not a link to the Greek optative (though the Greek optative might theoretically have formed as the result of some shortened stem verb plus a form of ειμι or ιημι, though the root verb form used is a stem, rather than a variable and conjugated form or an infinitive - PERHAPS it was a participle form with ιημι).

As for only Greek and Sanskrit having an optative, I think you are confusing constructions that have optative meaning with an optative conjugation. All languages can identify a sentence as an optative sentence in certain instances in terms of meaning or intent, but hardly any languages have a special construction to set off an optative (most of them being agglutinative languages with no proper "conjugations" to speak of, unlike Indo-European and Semitic languages that conjugate verbs). The only conjugated languages with an optative conjugation that I am aware of are Greek and Sanskrit, and both drop out of the language in a hurry so that level of distinction was obviously not needed in conversation. Romanian doesn't have an optative conjugation in that sense.

I read the Smyth paragraph you cited and now understand what he's saying - he's talking about the vowel length and accent being identical rather than the endings. I still find him maddening generally in that he makes authoritative statements but we don't get citations to longer works - ever.