If it had been someone like Benoot or Vanmarcke in that break with vdPoel, someone who we know has a track record of losing 2-up sprints in classics or at the end of long breakaway stages, you'd probably be right, and they would probably have gone on to lose that sprint. But Asgreen, to the extent that his sprinting was a known factor, it turns out actually had a pretty good past record in those situations, to the extent that in retrospect, vdP's win was not the gimme we all, and he, and everyone except Quickstep, thought.
You thought it was foolish of Asgreen to work with vdP at the time, and that was a fair assessment with the information you had at the time. It has since come to light that Asgreen can win match-sprints at the end of long races, not only yesterday, but in other previous races too, something Asgreen and Quickstep knew but the rest of us either didn't, or dismissed.
What you haven't acknowledged is the possibility that, if Asgreen had sat on and not taken a turn, vdP might not have ploughed on 100% all the way to Oudenaarde. vdP was the favourite in the sprint with Asgreen. If the group with GvA, Turgis, Senechal and vdP's teammate Vermeersch, or whatever combo of Alaphilippe, vanBaarle and Bahrain was in the moveable feast that was G2 on the road at any given time in the final 40km, come back to them, vdP is still the favourite in that sprint and Asgreen drops to 4th or 5th favourite.
Again; working with vdP, Asgreen's worst possible result was 2nd. If they got caught by the group behind, Asgreen's worst possible result drops down to 5th or 6th or 8th or 10th, and his chances of 1st are also significantly reduced, and that's even if we accept (which Asgreen didn't) that vdP is going to win that sprint 9 times out of 10.