An interesting contribution Adam. I am not a trained historian (beyond A-level anyway) but have spent much of my professional life collecting, analysing, evaluating and presenting information. Primary sources, such as are being discussed here, are clearly an essential part of historical research. However, the fun begins when primary sources disagree, as they often do.
A good example which keeps coming up is the effectiveness of the Allied fighter-bombers in knocking out tanks in Normandy and beyond. You will find many first-hand reports from fighter-bomber pilots describing in detail the great execution they wrought. The problem is that Operational Research teams, who crawled all over the battlefields shortly afterwards trying to find out what knocked out the tanks, found that very few could be attributed to aircraft: the pilots had basically been mistaken, for a variety of reasons.
Personally I welcome the publication of raw primary sources as such books save authors like me a lot of slogging through files, although they should perhaps come with a health warning to the general reader that the contents only represent what the originators of the documents thought at the time, and may have been contradicted by later research.
Tony Williams: Military gun and ammunition website and discussion forum