I know that prior to the first reports of ethnic cleansing, the Breadline massacre had occurred which really started to hint to the international community that something was wrong. I also know that ethnic tensions had existed ever since the Ottomans left the area.
Looking backwards, it probably was predictable, but with hindsight almost anything is. Ethnic tensions were present, but they were seldom a key issue, until they became acute. As ethnic history of the area is as much a history of cooperation than conflict, or even more so, existing tensions as a predictor of ethnic cleansing really only works in hindsight. In late Yugoslav decades, ethnic relations were often a point of concern, but seldom were they perceived as a immediate and potentially dangerous issue. Outside Yugoslavia, an image of Yugoslavia as an successful multi-ethnic state (if not at the same time as of successful socialist one) was quite prominent, so analysts that predicted ethnic violence, if there were any, really came into their own only after the fact. Finally, it is worth noting that ethnic violence was more of a consequence of how Yugoslavia was falling apart, of nationalism that replaced socialism as an legitimation principle, than the cause of the break up.
With all that being said, and keeping in mind that predictions with hindsight are easily made, there were some clues to possible ethnic violence. War against "Turks" (Ottoman Empire) were an important common point of Yugoslav history, at least for two dominant nations, Serbs and Croats, more so for the first, but important for the second as well. "Turks" here is an important reference, because in most popular speech it was used as an ethnic marker, rather than political-imperial one of "Ottomans". Also, by common popular connections, "Turk" and "Turkish" were often identified with both muslim and Muslim. Within standard Yugoslav notation, lower-case muslim denoted a religious affiliation, while capital case denoted the people/nation (later named Bosniaks). Possible implication of that Turk-Muslim-Bosniak identification was that they were a foreign, introduced element and therefore also a prime target for cleansing. But again, it would be very difficult to make that argument for the danger of ethnic violence, before it actually occurred.
Another possible predictor would be an emerging ethno-nationalist legitimation principle, gradually ever more important during the 1980s. If ethnic majority would be an argument for new borders in the case of the break-up (replacing those set up in 1943, or rather 1945), than significantly mixed areas had the potential to become sites of ethnic cleansing. However, both predictions would first require a prediction of Yugoslav break-up, something that was not often considered. Than, it would require that borders were questioned, while if there even was any thought about the break-up, preservation of AVNOJ borders was mostly assumed.