EDIT: loose dogs that chase are one reason why it might be SAFER to have a higher top speed. I clocked the one on my street doing 35+ KMH and he gave up when I hit 40 KMH. OTOH, I went much slower, to save my battery for the race, on the way back and he only came to the end of the driveway and barked a few times.
I've had my bike to 56 KMH (35 MPH), both downhill gravity powered, and down a short hill with my 20s1p batteries half depleted and partly damaged. 52 KMH or 32 MPH have been easy to reach on flat.
In theory, the calcs and simulators indicate I SHOULD be able to get 65 KMH (40+MPH) on flat with 20s1p charged to 84v open circuit. I'd like to see that just to prove the calcs, or even 60 KMH (37.5 MPH).
I don't NEED this speed, but it's fun to try. I'm more interested in range and hill climbing with best efficiency. With a front XLyte 404 and 20 amp controller on 19.1 inch wheel my top speed with 5s4p should be about 27 KMH or 17 MPH. I've tried 5s1p before and thought it would be laughable, but it was surprisingly OK for the low nominal voltage of 18.5v.
IMO, 17 MPH should be plenty for my "pedestrian range extending" rides on bike paths and such. 5s4p, or perhaps even 5s8p could be really useful for keeping my batteries in good shape and I'd be worrying much less about voltage drop and end of ride LVC.
I keep thinking I need a system with FETs that can switch me from 5s8p to 10s4p or 20s2p on the fly so I can have fun at higher voltages or extend range and cell life at lower voltages. Maybe I'll build this.