According the Puranas, Buddha's birth was predicted - to happen later. So there is no contradiction, rather there is agreement in the timing of his birth to be around 2600 years ago (IIRC).
The Puranas were an oral tradition before they were set in writing, and since the system in India was to write on palm leaves, which had a rather short life span, the four Vedas and their corollaries such as the Upanishads, Puranas and so on were copied and recopied continually, sometimes with new commentaries or additional material added. That is probably why some historians consider them of more recent origan; that and the fact that historians generally have a set idea of history and like things to fit into their theories rather than be open to different time scales.
That's where I'm seeing the issue. You can't document an oral tradition prior to its earliest verifiable transmission, and prior to that there's no verifiable way of knowing what happened to the tradition before it was committed to writing. That doesn't mean the tradition thereby isn't true--it may or may not be true in a given case--but it means a historian following the empirical method can't affirm an oral tradition as documented history until it's substantiated by some corroborating evidence. I agree with you on some historians having a set idea of history, but that isn't the issue here for me. The issue for me is that the documents we have don't date back as far as the events they're referring to, so they're not primary sources in the historical sense.