But since morphemes of Greek and Sanskrit can be shown to be related by applying regular sound changes, this is not even relevant.
For instance, Sanskrit ending for 3rd prs sg present is "ti", and Ancient Greek has a zero morpheme there. How to explain that? Was there a stage in AG when final i-es weren't allowed but final t-es were?
I doubt that
I haven't looked much into the Liburnian language, I just assumed what's written on the Croatian Wikipedia about it to be the mainstream science. Besides, explaining the apparent Anatolian loanwords in the Chakavian dialect by Liburnian being a late Anatolian language doesn't appear particularly unreasonable.
h2en means "there" in PIE. There is a pronominal root h1e, so its nom ntr may be h1en, "that". Are both roots related? I do not know.
*h2en is the source of the English preposition "on". In PIE, it was a postposition meaning "by" or "near". PIE postpositions obviously easily become prefixes in daughter languages.
What is "dheh2"? I looked up the PIE root for "flow" and found only srew and plew. There is a root dheu that means "flow", attested in the Germanic languages (one of the subfamilies of IE I am quite knowledgeable). I do not know if it is a PIE verb.
*dheh2 is the source of the root *danu (whence, for instance, Danubed) in the Old European Hydronimy. That is, *danu=*dheh2+*nu, "that which flows" (or, more accurately, "that which is flown"). The Latin word "fons", I think, is also said to come from *dheh2 (although ablaut doesn't happen regularly in the participles).
How exactly do you explain the ending of Andautonia? When om is meant to be the nom ntr marker, there can be no formative suffix behind it.
"-onia" was a common ending in Vulgar Latin, there is no reason to think it was the native one.