It's mostly due to aerodynamic differences in rider position for the various bikes you selected. Choose the same bike, same weight, same rider position for the different tires and you'll find out how much difference the tires make. It's significant, but not nearly as extreme as the differences in aerodynamics.
This is relevant to us because most of us need tougher tires than recreational or sport riders would use.
It's plausible to put an aggressively aero riding position on an e-bike, but when most of your power isn't from pedals, that can get foolishly uncomfortable in a hurry.
For what it's worth, 19 out of 20 of us here use bikes that either correspond to "roadster" settings, or are draggier yet.
This is starkly illustrated in the difference in efficiency between two of my builds when I had identical motor/controller/battery setups in them at different times and places.
70 lb DB mountainbike w/ full suspension, using a Leafbike 1500W 4T with a single Greenway 46.8V 15.6AH LiIon battery pack, Phaserunner controller, Schwalbe Marathon Plus Tour 26x1.75" rear tire, WTB Tyrannoraptor 25.5x2.5" front tire, CA3 restricted to 750W, torque-sensing PAS through FAG torque sensor(yes, it's GAY because of it's long 127mm BB spindle length, but it is what was available and works...):
-30 mph, moderate pedaling effort: ~17-20 Wh/mile, 30-40 miles range
-30 mph, no pedaling effort, throttle only: ~28-35 Wh/mile, 20-25 miles range
-37 mph top speed with hard pedaling effort and 750W max power input to motor
-30 mph top speed with no pedaling effort, throttle only)
-25 mph top speed pedal only, cogging torque losses cancelled by controller
82 lb KMX-based velomobile w/ front suspension, using a Leafbike 1500W 4T with a single Greenway 46.8V 15.6AH LiIon battery pack, Phaserunner controller, Schwalbe Marathon Plus tour 26x1.75" rear tire, Schwalbe Marathon Greenguard 20x1.5" front tires, CA3 restricted to 1500W, cadence-sensing PAS through Sempu(which never worked correctly as a torque sensor, but its 115mm spindle length played nice with my front mech):
-30 mph, moderate pedaling effort: ~7-8 Wh/mile, 85-100 miles range
-30 mph, no pedaling effort(throttle only): ~12-14 Wh/mile, 50-60 miles range
-46 mph top speed with hard pedaling effort and ~1500W max power input to motor(although drew ~600-700W at this speed)
-43 mph top speed with no pedaling effort(throttle only)
-35 mph top speed pedal only, cogging torque losses present and battery shut off
When I added a second Greenway pack in parallel for the velomobile and increased weight to 91 lbs and power eventually to 3 kW, it was able to get up to 50 mph with hard pedaling and had no problem cruising at a top speed of roughly 45 mph on throttle-only while drawing 1 kW or so. This top speed was very obviously voltage-limited. The extra battery pack in parallel kept the voltage drop down and allowed a small gain in top speed as a result. Range was a consistent 150-200 miles at cruising speeds of 30-35 mph with light to moderate pedaling effort.
I briefly tried the aforementioned 4T 26" motor in the velo at 10kW with hubsink, ferrofluid, an ASI BAC4000, and 46.8V 33.6AH Molicel P42A pack. Top speed was unchanged vs the 3 kW peak Greenway pack, but acceleration was very violent.
I ordered my 2nd 1500W Leafbike motor in a 20" wheel instead of a 26", and as a 3T instead of a 4T, but run at 10kW and using that same 46.8V 33.6AH Molicel P42A pack and the hubsink transferred over from the 26" wheel. I used a 20x1.5" Schwalbe Marathon Plus in each wheel. It was dangerous to accelerate at full power with 0-30 mph in about 2.5 seconds and top speed was still 50 mph despite the smaller drive wheel. The smaller drive wheel also caused ground clearance issues with the shell(I added a rear suspension). The rear tire quickly lost its tread. I then tried the rear wheel with an upgraded rim and Mitas MC2 16x2.25" DOT tire, and eventually doing this to the two front wheels as well, which gave slightly more ground clearance, about 1/2" more. It wasn't enough. I removed the body shell soon after. It was originally designed with the 26" wheel in mind, and the 20" bike or 16" moto wheel out back, swinging on a suspension, caused issues between the shell and rear wheel over deep potholes, as well as unwanted scraping f the tail with the ground that didn't happen with a 26" wheel.
With a reconfigured 72V 25.2AH Molicel P42A pack using all the same cells and then some more, I ran it with the body taken off. It was more than a bit nuts thanks to losing about 20 lbs and having more voltage. I never topped it out like that because with the body shell off the trike it felt a LOT more dangerous over 40 mph or so than with the shell. With the shell, it felt safe, even though the protection provided was mostly illusory. Without the shell, it makes me very confident of my impending death, or at the very least, my wallet and keys falling out of my pockets and landing in the street. I'm confident as a bare naked trike, it would probably have done at least 60 mph like this, but that aero drag at such a speed would have quickly have made that motor overheat.
I didn't have access to a wind tunnel or use any CFD software when I designed the shell, so a lot of room for improvement exists regarding the drag of my velomobile. Which is why it's apart and why I'm working on a new body shell for it, aside from the roll cage and AWD upgrade, and new wheels/tires/full suspension/hydraulic brakes/ect. Because of aerodynamic improvements, I expect when ready, energy consumption will greatly decrease over the previous iteration, in spite of being prepped for 60+ mph cruising, having a hubmotor in each wheel, having DOT wheels/tires, and weighing upwards of 120 lbs. And I still expect, using nothing but my own legs, to be able to maintain 25+ mph cruising speeds and sprint to 35+ mph top speeds with the controllers being used to have the cogging torque cancelled, and maybe bicycle-like pedaling speeds with a disabled battery and powering through the cogging torque losses of all three motors. Using all three motors including a Leafbike 1500W 3T rear powered by an ASI BAC4000 with field weakening, x2 Grin All-Axle 3T up front each powered by a Phaserunner, each motor in a 16x1.5" DOT rim with 16x2.25" Mitas MC2 tire, all run by a 72V 25.2AH pack of Molicel P42A LiIon batteries, I expect that it will out-accelerate most new cars to 60 mph and reach over 80 mph, this top speed again voltage limited, because the damned thing might have 20 kW under that setup, or even more when I upgrade the front controllers to something better. Range at 30 mph could end up at over 300 miles and range at 60 mph could still be 100+, regardless of whether I'm pedaling or not. In most jurisdictions, I can only legally operate it at 28 mph/750W of assist, so real usable range would probably be ICE car-like.
The mountainbike will get a set of fairings, but I don't expect to cut drag by more than 1/3. That is still valuable, as it means pedaling provides an increasing proportion of the vehicle's thrust at a given cruising speed in torque-assist PAS mode as efficiency increases and power requirements decrease. So the new range will be more than inversely proportional to drag as a result, range and drag relative to each other before and after the modifications. Range at 30 mph with pedaling might double for the mountainbike on a similar sized battery thanks to said faring, pedal input effort being a constant quantity between the two iterations. But my new battery for this bike will be slightly smaller than the Greenway pack. I plan to have a continuous power of 1,500W and will set that as my limit via CA3. I expect 40-ish mph throttle only with the aero improved and 1500W available. Maybe 50 miles range at 30 mph with light pedaling effort on a smaller 46.8V 13.2AH custom pack, again once the aero work is done.
We shall see.
