I can understand, given the nature of these projects, that it might be necessary to batch process the result data. But it seems to me, given the probably pseudo-random arrival of results from around the world, it might be more efficient to have the server either

a) Process individual results in their entirety as they are received
b) Split out the stats and update those right away, while saving off the real project results for batch processing.

From a public relations perspective, I think folks are less likely to lose interest if they can see their results show up on the website right away.