It seems amazing that one could take an arbitrary sentence from one language, write it in a different script, and then read that script in a third language and get a coherent sentence.
Well, note that significant information was lost in the transition from Latin to Ventrisian syllables (arma -> ama; wirum -> wiru; Troiae -> toroia; primus -> pirimu; ab -> ap; oris -> ori), and even more was hallucinated in the transition from Ventris to Greek (ama -> halmai; kano -> kainoos; etc. etc.).
The point being made is that the standard for interpreting Greek from Linear B is so low that any set of symbols at all can be validly interpreted as a coherent Greek sentence.
As long as 75% of the words are viewed as proper nouns, of course.
You might also consider how a modern language with a very small syllable inventory, Japanese, customarily represents foreign words in its own syllabically-impoverished writing system.
For example, the English word "credit card", which is three syllables, is written in Japanese as ku-re-di-t-to-ka-a-do, which is eight. The Ventris hypothesis works in the other direction -- he believes that speakers of phonologically-rich Greek made do with a writing system that could only express much simpler syllables.
Though we have examples of that too; Antiochus of Macedon left an Akkadian inscription reading, as given in Empires of the Word [1], "I am An-ti-'u-ku-us [Antiochus], the great king, the legitimate king, the king of the world, king of E [Babylon], king of all countries, the caretakeer of the temples Esagila and Ezida, the first born of Si-lu-uk-ku [Seleucus], Ma-ak-ka-du-na-a-a [Macedonian], king of Babylon."
Contrast Antiochus' (or his scribes') concern for vowel length when he goes to the trouble of writing Ma-ak-ka-du-na-a-a with the reading of kainoos from ka-no.
