What I learned while researching my only (unfinished) historical fiction is that historians who read original sources pick and choose what they present in their own books and they synthesize to create a coherent story. The problem is, they often do their picking and choosing to support their personal theories. Sometimes I could do without their theories. Synthesizing is quite the art, nevertheless.
I think that's a valid point. In my own series, the amount of stuff I left on the "cutting room floor" really did outweigh the stuff I incorporated - and by a wide margin. Sometimes those arcane points make for some fine story telling, especially if one can pull it off. Of course, I don't write straight historical fiction, either, so take my opinion with a grain of salt.
It's probably controversial, but my view of historical fiction tends to be that the further one goes back in history the greater license you should be allowed to fill in the blanks, where those blanks are unknown.
Of course you can fill in the blanks that are unknown as long as it fits with what is known. That really isn't controversial. I am not sure what you mean by 'greater license' though. When I fill in blanks, I am particular about making it plausible and fit with the known facts. Of course, 'known facts' are not always easy to disentangle. Even original sources can vary considerably about what they think happened.
I do use original sources much more than historians because you are right that they interpret those to fit their own theory or out of laziness, just repeating what was said by previous historians without bothering to check the original source. Fortunately, it is quite possible to buy copies of many of the original sources of the period I write about. It would be a tad more difficult if I had to travel to Scotland every time I wanted to check a source.
ETA: I love doing the research and reading the sources or else I wouldn't write historical fiction. Even so, there are times when it definitely gets frustrating.