How do you detect revisionism by a past historian? What about good, old bias?

Revisionism isn't something to "detect." Revisionism is just a pejorative word for "something someone said that is not the standard story." Coming up with a new, better story is the goal of a lot of history.

The reason for the pejorative is that a lot of people associate "revisionism" with "history told in a new way with a lot of bias." But it's a dumb way to refer to it in my view. There are "orthodox" histories with lots of bias, there are "revisionist" histories with lots of bias — bias is the problem, not whether it is old or new.

As for detecting bias: if it's a time period and source base you know very well, it's pretty easy to see where things are biased. I've been thumbing through Margaret Truman's biography of her father and it is obviously super biased. Not just for the obvious reasons (she is her daughter, she portrays him in a universally positive light), but on things related to the atomic bomb it is just plain incorrect, citing things that never happened, putting things in the wrong order at times, etc. That's an extreme example, but if I gave you the chapters in question, could you spot the subtle biases, the additions and omissions? Probably not, unless you already knew the story really well.

You might think that a book with footnotes would make this harder to get away with, but it doesn't, unless you already know what's in the footnotes. I see stuff get cited all the time as evidence, but if you look at the original document yourself (with a ton of other information in your head), you can see that the historian in question didn't quite get it right. Sometimes that getting it wrong is just accidental (just "error"), but sometimes it is deliberate, a means of stretching the evidence to fit what they want it to be — that's the "bias" part.

Again, there's no magic way to do it. You detect it by being an expert and already knowing the source material and having the capacity and capability to check it all yourself. Sometimes you can sniff it out from the general narrative (e.g. strong heroes and villains are usually a form of a kind of bias — historical figures are just usually more complex than that), but if you're talking about anything subtle, it requires being an expert. Which doesn't make it any different than any other form of expertise (can you sniff out errors in a physics paper?), though people tend to underplay the importance of historical expertise because the "end product" looks a lot like "regular" storytelling.