We tested all different times and traffic levels but in general, Uber's algorithm is terrible at predicting traffic since they use their own internal maps system (as opposed to gmaps or waze).
Lyft has switched to an upfront pricing model too so it wouldn't surprise me if they are doing something similar. But I haven't seen any examples that are as bad as with Uber, where I'm constantly seeing big discrepancies like this.
We didn't test this with Express Pool but Pool in general is a different animal since the prices are so cheap to begin with, there's not a lot of variability that can come into play and affect the pricing.
Plus, Uber takes a huge loss on Pool and Express Pool especially when they first launch since they want to offer low prices to riders but they don't know the match % rates yet.