Thanks for all the responses. some really good points were made. To clarify where I'm coming from in asking this, allow me to present my understanding:
Jesus died around 30ad. He and his followers spoke Aramaic. The gospels were written in Greek at least 35 years later by unknown authors and there are some discrepancies between the 4 books, such as the exact day Jesus was crucified on.
The 4 authors are trying to emphasis different points in the story - John, for instance, wants to emphasis the parallels between the crucifixion itself and the moment of slaughter of the sacrifical passover lambs in the temple, so he times the whole story off that point. The synoptic authors have different priorities in their stories. We shouldn't assume that the authors have a post-enlightenment obsession with getting telling the facts accurately and the expense of the point of the story.
I would think that since Jesus was going around doing miracles, that someone would have wrote about it at the time, not decades later.
Bits probably were, but in a society that is less literate than ours writting things down isn't the first point of call. It is highly likely, however, that the central story of passion week was written down earlier, and then those accounts incorporated into the accounts of Luke etc.
I do agree that being able to compare thousands of ancient copies gives us a good idea of what the originals said - however this does not mean the the originals were 100% accurate historical accounts.
I disagree that the memories of people 2000 years ago were more accurate than ours. If someone knows of a study or something more tangible than conjecture, please post a link or reference.
It's not mostly about individual memories, but the collective memory of a story-telling culture. When the story is well known and oft repeated in a story-telling culture the collective memory preserves it's accuracy - if the story-teller deviates from the accepted story the community correct him pretty sharply (and sack him if he does it repeatedly).
Hence my example I mentioned before: at some time over 10,000 years ago the course of the Murray River stopped at a choke in its course. Eventually it broke through and that event is preserved in one of the stories of the local aborigines - preserving a (mythologised) account of an event that happened over 10,000 years ago and for which there is no current evidence available to a pre-scientific people. Storytelling can and does preserve stories for hundreds - even thousands - of years. 40 or 50 years is nothing.
We've lost the power of collective storytelling, but its power of preservation is well understood, which is why non-literate cultures turn all the important information they want to preserve into stories.
