"Two-Source" vs. "Two-Document" Hypothesis

So, let's see:

A GoogleFight comes out 15,100 to 815...

a Google Book Search finds 402 pages to 230...

and an ATLA search on EBSCO Host Research Database found 39 hits for "Two-Source Hypothesis" in any major field to 20 for "Two-Document Hypothesis".

I suppose that settles it.