Unsourced semantical and etymological connections that arose thanks to the project being very connected.
Regional Italian scopiglio／scopillo and Albanian shqopë, both “heather” and both ultimately from Latin scōpa(“broom, yarrow”). Only two descendants that maintain one of the original meanings of the Latin etymon.
Sicilian pisciuni(literally “big fish”), Sardinian pische de sa camba(literally “fish of the leg”) and Moroccan Arabic حوتة الرجل(ḥūtat er-rjal, literally “fish of the leg”), all “calf”.
“hundred dollars” not containing the stem of “hundred” in Romanesco Italian piotta, Cantonese 水(seoi2) and American English Benjamin.
Spanish trompo and Ancient Greek στρόμβος(strómbos), both “spinning top”. A pre-PIE root *(s)trómP-o- is reconstructible, and archeological evidence shows that spinning tops are very old. The Spanish word is claimed to be from trompa(“snout”), but that seems semantically difficult, while the Greek one is admitted to be from a pre-Greek substrate.
Tosk Albanianvapë(“sultriness”) and Italianafa(“sultriness”), both of unclear origin, and both regularly reconstructible to a form *āpʰā. The Albanian shift of *ā- > *ō- > T. va- ~ G. vo- is well known, cf. vatër. For the aspirated pʰ, early Paleo-Balkan languages likely aspirated their voiceless occlusives much more than modern Albanian does, cf. Greek loans lakër, lëpjetë, mokër ultimately loosing the aspiration. Cf. also the Illyrian king called Gentius by the Romans and Γένθιος by the Greeks.
Having spotted the Italian word alfalfa, assuming it was a regional, autochthonous term, Italic origin seemed evident: a form *alfalfā(literally “white-white”), from *alfos(“white”), whence Latin albus, and also attested in Umbrian. On a second thought, word-formation by duplication, while by no means rare, did not make much sense for this. Hence an even more plausible, and dare I say, undeniable *herfā alfā(literally “white herb”). After the loss of the word *herfa in the Romance dialect, the assimilation of the [erf-] with the following [-alf-] would be more than expected, since pretty much all dialects of Italy mess up pre-consonantal [r] and [l], I'd recommend one to check AIS 586. Turns out the word is actually an internationalism from Arabic.
Another stroke of genius was with the two Italian terms furbo(“cunning, wily, crafty”, adjective) and furfante(“scoundrel, rascal, villain”, noun), both quite clearly from an Italic verb *fūrf-ā-(“to trick, deceive”), the former through Latin as an *-o-adjective, and the latter through Italic as a substantivised *-ant-present participle. The Italic verb would be transparently akin to Latin fūr(“thief”). Turns out, Latin *fūrbus does not actually exist, as the word was borrowed from French in the 15th century.
[…] it is important to stress that this interpretation is not a poor, but a brilliant one. It conforms with difficult requirements, and the etymological equations on which it is based are almost evident. No doubt, this is the etymological method at its best. The proposed interpretation has only one fundamental weakness, namely that of being demonstrably wrong.
