Okay, this was a much better episode. Much like the season-premiere JATO rocket car episode, they revisit two of their most notable failures, cases where they didn't quite get things to work right in the time available, and try to figure out how to get them to work right this time.
Painting with explosives: I always liked this myth, and I really like the revisit. It's cool to see Adam and Jamie applying the experimental method, testing out designs and learning from their observations. It makes me feel all sciencey just watching it.
And "C4 jerky?"
This is a myth I have a vested interest in seeing proven viable, since in my novel
Only Superhuman there's a device called a "scene-painter grenade" that fools security cameras (which in the future could be embedded in pretty much any surface in a room) by coating the whole room in a spray of nanoparticles that "freeze" the incoming light and continue projecting light of the same color and intensity, so that any hidden camera lens sees a still image of the room until the particles wear out. So if Adam and Jamie could get this gimmick to work, it'd be proof of concept for my idea. (Although I hedged by asserting that the nanoparticles would flow together to form a solid coating first.)
I'm a little disappointed that Adam cheated and changed the parameters of what he was trying to do, since that left only one chance to get it right. And while the high-speed shot of the multicolored paint explosion was gorgeous, I kind of had a feeling that the resultant "art" would be just the kind of dull brown or gray you get when you mix too many colors together. And that's pretty much exactly what happened.
I also wasn't very satisfied with Jamie's rig, which seemed a few steps backward from the Flower of Death he got moderately good results with last time. I might've liked to see a further refinement of that principle. Now, his idea of the sphere with nozzles had potential, but why did he limit it to only four nozzles? If it had had many nozzles pointing in all directions, that would probably have worked better. Although I suppose too many holes would've meant too little blast containment. But there's got to be a better balance.
Bifurcated boat: This was more of a conventional revisit, not just trying to perfect something they didn't manage to pull off before, but responding to fan criticisms about their methodology. Still, it was worth revisiting. The result of the test was interesting -- the channel marker breaking instead of the boat.
But that leaves me confused about the ultimate result. They got the split at high speed, but only by using a metal pole. So how come the supposedly wooden pole in the original photo was intact?