From what I can gather, OW5 was built using Zooniverse-common subject presentation and tracking software elements, which are designed around random subject selection. I haven't worked a lot of Zooniverse projects, but randomly presented subjects don't seem to be a problem for most of them. In fact, for many things it's probably a desired method as a way of minimizing observer bias in sequential image evaluations.
The details are really down in the weeds, but that developer discussion hints at the problems that they're having trying to shoehorn the existing software into a date-sequential mode. But they are (or at least were) trying, so our complaints have been heard!
It is a shame, though, they didn't see fit to adapt the OW3 workflow to a newer software system, if that was the problem.
Whaling, BTW, was built around something called Scribe, which is an entirely different system. The random presentation of transcriptions probably draws from the same thinking, though.