^ I keep hearing this, and I take it to heart. But I don't like any of the noclean fluxes I have tried. They are great, until you actually ever want to clean the board. Add alcohol, and watch the transparent "polymer" residue turn into schegma. Most of the time I don't clean off the residue, and I still can't stand this.
This maybe the reason the only flux that is mil-spec is rosin flux?
I would really like to find a rosin flux like what Indium uses in their solder. They say it's a purified rosin flux. It leaves something like regular rosin flux residue, but with a lot of the yellow/brown color removed. I'm using regular liquid RA, and lots of it. My bevel tip is normally operating with a crust of brown flux residue around the sides.
then again at 19:00 he adds more flux and reheats with the soldering iron what looks like a perfectly soldered part.
He even says this is not necessary while he's doing it. And it might look the same. But I can see why he might have done it. He flowed flux core solder onto the pads with an iron, using up some of the flux. Then he reflowed with hot air. So the solder could be a little bit crusty, theoretically. If he were doing a leadless IC like that, he probably would have added more flux BEFORE reflowing. He might have felt like something was missing, after the fact.
FTR, I am definitely not a perfectionist. Most of my SMD joints would make you laugh. Straightness doesn't matter, and anything from a concave filet to a grossly spherical bead is fine, so long as it's shiny and obviously flowed to both the lead and the pad. IME, you don't need a perfectly controlled amount of solder to be able to see that.