What other way to learn is there than trying to analyze the reasons though? It's not like Joshua is using some black magic to make those decisions. They are logical and deductible, even if tricky to do so. You can disagree with Dr Dave's opinion about what the correct shot was, but I don't think criticizing him for stating his opinion is very fruitful.
As for the actual topic of the video, IMO it all boils down to what % Filler felt like he would make contact with the OB with the 3-railer. If he deemed the percentage high enough, then he might've felt that the overall probabilty of winning the frame is higher with that route giving better leaves on average with the speed he hit it with.
For example, if I was faced with that situation, aiming to maximize odds of hitting the ball, I'd estimate my 3-railer to make contact maybe 40%, and the 1-railer like 98%. I'd probably think that the 3-railer gave slightly better leaves on average, but nothing near enough to justify choosing a 40% path over a 98% path. It's not like I'd be doing this math while playing, but just the general feeling of "almost guaranteed hit with slightly worse leaves on average" vs. "less than 50% to hit with slightly better leaves on average" makes me select the first option as the correct one for me.
As for what % Joshua would hit the 3-railer, we'll never know for sure. But seeing that he selected it, I suspect he estimated it to be high enough for it to be worth it. Even at 80% contact (just guessing at a figure for the sake of an example) it might flip the 3-railer into giving better overall odds than the 1-railer. Judging this more precisely would require analyzing the leave odds of both routes more deliberately. My analysis is simply based on the general feeling of the 3-railer giving somewhat better leaves on average with the speed Joshua hit it with compared to the 1-railer, most notably due to the cue ball staying near the end rail with high likelihood.
