As promised I did look into the subject. I might add two things at the very beginning. First, I'm in no way an expert either on Transylvania or this period. What I know comes from my general knowledge of Hungarian history and some reading on the subject for the purpose of writing this blog. Second, my knowledge comes from Hungarian sources, which may color the thesis that the Latinized Dacians couldn't have survived unnoticed for about a thousand years after the fall of Dacia to the "barbarians" and hence modern-day Romanians are not the direct descendants of the Romanized inhabitants of today's Transylvania.
In 1987 a monumental three-volume history of Transylvania appeared under the imprint of the publishing house of the Hungarian National Academy: Erdély története három kötetben. The editors– László Makkai, András Mócsy, and Zoltán Szász–were the foremost historians in the field. Each longer part or chapter was written by a recognized expert. Endre Tóth wrote about Dacia, István Bóna about the period of migrations (271-896), and László Makkai about the early Hungarian rule (896-1526). Just to give you an idea of how thorough the treatment is, these three chapters take up more than 350 pages. The three volumes are a hefty 2,000 pages. Therefore, I must drastically abbreviate the story and concentrate only on the most important themes.
A sidenote: when this book came out Ceausescu's Romania was outraged. Romanian historiography in those days was blatant, primitive propaganda. To give a taste of the situation then, Ceausescu's own brother was one of the most celebrated "historians" of the country.
Let me start with Endre Tóth's description of the fall of Dacia. Archeological findings of Roman cemeteries and villages prior to 270 are plentiful; however not one Roman site was found that could be dated after the collapse of Dacia. According to Tóth some Roman coins and Christian articles were found past this date but they are not enough to prove continuity.
The Roman province of Dacia was first overrun by the Visigoths; under their rule today's Transylvania became almost uninhabited. Soon enough another Germanic tribe, the Gepids, appeared on the scene; they originally came from the area of the Vistula (in today's Poland) and moved into the region before the Huns showed up from Central Asia around 370. Eventually the Gepids became the favorite Hun vassals, and they took part in some of Attila's most famous battles. Therefore they received Attila's permission to settle in the Carpathian basin. After Attila's death in 457 the Gepids turned against the Huns and established their own kingdom that almost perfectly maps the borders of today's Transylvania. This date can be pretty well corroborated by archaeological finds.
The Gepids brought to Transylvania a surprisingly high level of cultural attainment. However, after about one hundred years a new group arrived from Central Asia, the Avars, a multilingual tribal society headed by a central ruler called the khagan. The Avars, just as later the Hungarians, were being pushed from farther east, and for a number of years they tried to escape the attack of other Asiatic tribes by somehow getting beyond the Carpathian mountains.
And here let me elaborate a bit because I found the description of the Avars' difficulty getting into the Carpathian basin intriguing. The Gepids apparently made sure that there was an 80-100 km wide area in the high mountains that was left untouched, heavily forested, with no mountain passes. That was apparently their equivalent of the Great Wall of China. The Avars tried several times but simply couldn't penetrate the Northern or Eastern Carpathians. So they shifted tactics and moved southward. First they attacked the people living in Moesia (today part of Serbia, Bulgaria, and Romania), later moved a bit westward into Sirmium (Srem in Serbian, Szerémség in Hungarian) and eventually managed to get into the Carpathian basin relatively easily from the south at the end of the sixth century. They defeated the Gepids who were either killed or absorbed.
The Avars were not as clever about the country's defenses as the Gepids and soon enough Slavic tribes were able to penetrate the narrow mountain passes. In the sixth century Slavic tribes moved toward Moldavia-Wallachia, the two Romanian provinces that eventually constituted modern Romania (minus Transylvania and Bessarabia). By the seventh century there are archaeological finds attesting to the presence of Slavs in Transylvania as well. Slavic presence is also evident from some of the geographic names: Bystra/Beszterce (from Slavic "fast"), Krasna/Kraszna (from Slavic "red"), Congrád (from "cerni grad" meaning black castle).
Finally, a few more words about the society the Hungarian conquerors found in 895-896. By that time Slavic tribes had moved into the area in such great numbers that the Avars for all practical purposes had disappeared. And then by the middle of the thirteenth century Hungarians were the dominant force, inhabiting practically all of Transylvania with the exception of mountains higher than 2,500 feet.
All in all, it is hard to believe that Romanized Dacians could have survived the Visigoths, Gepids, Huns, Avars, Slavs, and Hungarians. To my mind an "outlaw society" (and I don't mean this disparagingly but rather to indicate a society outside of the mores of the dominant group–i.e., Dacians hiding out in the mountains) is at best marginalized by historical forces. But, to continue the wild west metaphor, there's no smoking gun to squelch the Dacian theory. There's always another side to the story. And that's what makes life, especially for a historian, so fascinating.