It's not right to say "bWAR assumes pitchers have control over the outcome of every ball hit into play". First of all, every outcome, including home runs, strikeouts, walks, involves some luck. Great pitches get hit. Bad pitches get missed. Calls get missed. Balls miss being a home run or a fair ball by an inch. Pitchers exert only partial control over every outcome. The best you can say is that pitchers exert more control over some of them (home runs, Ks, BBs, infield pop ups), and less over others (fieldable balls in play).
"Control" is an unfortunate word choice here, I think it gives people some flawed ideas about what's going on. I think it's better to speak in terms of probabilities. By pitching well, a pitcher can reduce the probabilities of bad outcomes happening, and increase the probabilites of good outcomes happening. This applies across all kinds of outcomes, and not just the FIP outcomes. But what those probabilities are are only partly influenced by the pitcher's performance. They're also influenced by other things. The probability of a hit on a ball in play is influenced by the quality of the defense, and the park dimensions. The probability of a ball or a strike is influenced by the catcher's framing ability, which umpire is at home plate, and how well the batter is doing that day. Home runs are influenced by park dimensions, temperature, humidity, and wind strength and direction. And who is hitting.
The Baseball Reference approach to pitcher WAR is to assess how much influence some of those other things likely had on the pitcher's run total, and adjust the pitcher's allowed run total from there. The quality of the defense behind the pitcher is one of the things they adjust for. This approach considers 100% of a pitcher's outcomes.
The FanGraphs approach to pitcher WAR is to disregard the 63% (on average) of a pitcher's outcomes that are more greatly influenced by other (non-pitching) factors, and to only regard the 37% (on average) of outcomes that are less influenced by other factors. Like Baseball Reference, they also make some adjustments that consider the influence of some of those other factors. But I find this overall approach to be inferior. It just disregards too much.
Here's a study I did of a case where these differing approaches made a huge difference:
