The Θ sound in Greek is certainly a remain of the pre-Greek languages.
I'm tend to believe that such a language was of the anatolian type. If not, it was a group of its own with heavy anatolian influence.
gcle2003 comment about Trojan is a good point. The only speculations we can do about Trojan language are based on Homers namings of people and the fact that geographically the Trojans were surrounded by certain people. Some say based on the names that Trojan could be some cousin people to the Greeks with a related language. Others say that Homer converted the original Trojan names to Greek. My input on that is that Greeks converted non Greek names to be more compatible with Greek speech e.g Navouchodonosoras is the Greek version of Nebuchadnezzar. However, unlike some Trojan names, Navouchodonosoras has absolutely no meaning in Greek.
The only Trojan record I've seen so far is from a man called Gaios (masculine form of Gaia) in Crete, claiming he is Trojan, using the local Cretan Greek (not eteocretan ofcourse) dialect:
http://epigraphy.packhum.org/inscriptions/main?url=oi%3Fikey%3D200820%26region%3D7%26subregion%3D26%26bookid%3D293%26caller%3Dsearch%26start%3D370%26end%3D374I do not believe Greek was an imported language but rather a result of a clash of languages in the helladic area, were neolithic inhabitants met new IE invaders. Based on that, whatever linguistic group gave basis to the new language, the same group could have some connection to the Trojan languages as well.
What troubles me is that Θ appears in Thracian writting but, from my understanding Thracians pronounced it a Z not Θ like english "Thermal" or Spannish Zaragosa.
Now, the same Θ sound appears in words like Thalassa which is clearly a Luwian word meaning "our salt" or something showing possession of the alas (salt, Luw. Alat).
However, the Luwian word for sea is "alasammis", so the Θ sound is a later addition founded in Greek. That brings some damn trouble and that's why i might suspect that the pre-Greek languages could be a group of their own with anatolian influence.
Something else that troubles me as well are combinations of consonants that are not that common in other IE languages nor other known major language groups e.g ΦΘ- like Phthora, ΣΦ- like Sphyrida or Mac. Sphryraena, Δζ- like Laconian Doric Dzini, ΒΓ- like evgenis (eugenis).