I rode two different ebikes. Both are 250w with similar sizes batteries. One had a mid drive and the other has a rear hub drive.

The mid drive seems to have significantly more in terms of efficiency.

It also has what imo is a less efficient gear system. The hub drive bike has a Shimano GRx derailleur system while the mid drive bike has a belt drive and a nuvinci hub.

I think the reason for this is that a hub motor inherently is less efficient than a mid drive because it has less favorable conditions. Hub motors typically have 5:1 gear ratios. Mid drives have 18:1. So the mid drive can more easily be in the optimal power band while a hub drive motor spends more time and power in non-optimal situations like starting from a complete stop.

One item that jumps out for me is the NuVinci hub, which from what I gather is less efficient than geared internal hubs, perhaps in part being CVT.

My mid drive e-bike has a standard 1x12 drivetrain and there's little evident loss. Do not have a hub-drive e-bike to compare against, but do have a belt-drive Shimano Alfine 11 commuter bike that is wonderfully smooth, silent and seemingly efficient. I'll note the Alfine hub costs several times what the NuVinci seems to go for, a Rohloff is even more that it. That's the gold standard.

Geared hubs run heavy—ease of maintenance vs. a standard drivetrain being a reward, along with the option of going belt vs. chain.

Lots of variables.

An internally geared hub is less efficient in general, but a (more expensive) mid drive makes it a non-issue. It's designed for ease of use, and less maintenance than a rear derailleur. Being a CVT with the NuVinci hub making it less efficient than geared IGH's is something I've never heard of. It's considered a premium IGH, like a Rohloff. I don't think it has the same gear range, but it's a great IGH nonetheless. I had it on my Xootr Swift. My only gripe was the weight, because it made my folding bike a boat anchor. An electric bike negates that issue.

As to efficiency in general, mid drives are definitely more efficient. Up until recently they were also hella' expensive. The pricing for them has now become reasonable. An IGH isn't always necessary either. I'd opt for a non-IGH mid drive and enjoy if that's your preference.

Me, I'm still waiting for my Lectric XP Lite JW Edition to ship. Rear hub, single speed, BMX style, 14+ AH battery... All the fixins for a simple, enjoyable ride.

How do you determine "efficiency"?
Do the batteries have the same capacity?
Do the motors have similar power output?
Does the mid-drive motored ebike travel longer distance with the same battery drain?
Do either ebike have throttle or only pedal assist?

So I have the Enviolo (aka NuVinci) IGH. It is a CVT, which means you lose from 4 to 10 percent. It is also significantly heavier than a 1x12 drive chain. However...
1. Perfect chainline like any SS, chains last longer as do sprocket/chain ring - YMMV
2. The Enviolo Electrique sets the near infinite gear ratio to exactly the cadence you set it to. So if you tell it to give you a 70 or 80, or 90... it will go up or down to maintain that cadence at the speed you are going. Pedal faster and it upshifts and you go faster. Pedal slower it downshifts.

So you get efficiencies from always being at the right cadence or close to it. I can change the cadence while riding, making it easier or harder to pedal, depending on conditions. Handy if riding with a group and you want to go faster but not spin like crazy. All in all it is slightly less average distance out of a battery, say 5%? Hard to pin down because how much I can contribute changes, so the distance changes.

Mine is a 40:20 ratio, so I get from 40:10 to 40:40 with the 380% gear spread. For me that means 5 mph up a climb and a top speed of about 36 assuming I don't go lower than 60 and more than 120 cadence. a 2:1 ratio is the lowest they recommend

Bike is powered by a CYC Photon, I rarely use more than 300 to 500w, although it will put out 1500w and the Enviolo HD will handle it plus me pedaling

How long the battery lasts is a real measurement most of us can appreciated. . This is affected by weight, tires, wind, hills, ride speed and how hard the rider wants to pedal, so everyone's numbers vary, I've found that I tend to pedal with the same effort on my BBS02B and hubmotor bikes, and it's really kimd of light work,. I'll burn the same watt-hours on the same route. Typically 10 WH/mile, but often a little less.

With my torque sensor mid motor bike, half the power of the BBS02B, it requires the rider to put in some work before the torque sensor responds,I could set it up for less force, but I prefer what I have. Result is I'm at 8WH/mile. Some of that is narrower tires and lighter weight, but a lot of it comes from the ride (me).

The point I was trying to make was that the mid drive is burdened by a less efficient drivetrain. The belt and the enviolo hub are less efficient than a chain and derailleur. The chain and derailleur setup is on a hub motor.

and yet, the mid drive is still more efficient and has more range than the chain-derailleur-hub motor.

this makes me question the claims of regen braking and hub motors. Because a mid drive makes more efficient use of the battery to begin with that even accounting for hub motors and regen it is less.

This seems akin to calculating a car's crank versus wheel horsepower. IDK how you can do so definitively whiteout testing apparatus and common methodology.

I was sold by the idea of less maintenance on rear hub motor.
Top end speed and accelerating power were less important in my buying decision.
As an aging cyclist, I want a boost in hills/mountains and for longer rides, 40+miles.
Class 1 to EU 20kph is what I ride.
Mid-drive felt more like a motorcycle than a bicycle.
ymmv

I mean there is always "physics" and *math*, There are very well known formulas and ways of calculating it without testing apparatus and the common methodology is physics and math.

On the design side sure, but empirical data are still required on the physical side.

The motor manufacturer usually publishes the information you need from their testing. That is really all you need.

How about dual hub? I’m eying the Eahora Romeo 2 now. It has a dual hub motor on both

there is more to it than this but a motor has an idea rpm where it is the most efficient. a mid drive is designed for that to be when you are peddling at around 80 rpms. a hub drive is only going to meet that at its max rpm. so if you kee a steady cadence the mid-drive is going to be very efficient.

My Fazua Ride60 mid drive has about 80% efficiency according to the manufacturer. It's on a Cervelo Rouvida 1x, 12 speed. It has three power profiles that can each be customized with a maximum power after a specific rider power input and and an acceleration rate for each profile. The battery is 430Wh. Max continuous power is 250 watts with up to 450 very briefly. How long the battery last depends on the needed assist. I've ridden 103 miles with 8740 feet of climbing and still had 9% left, but the easier parts of the ride used a profile that only added 80 watts after I input 120. It consumed less than 1% per mile. Climbing a 10% grade can consume up to 4% per mile, if the profile allows 200 watts. I only weigh around 135, so it doesn't take as much power when climbing.

Area 13 and TailHappyTV have evaluated this bike on YouTube,and one or both usually provides mileage values which might help. I'd guess 15 - 20 miles ridden fast.

Any idea how many wh per mile you get ? Just bought a power meter to monitor usage depending on terrain. On a very old Bionx 250w rear hub drive, I use 5wh per mile on a flat route at power assist level 2 (50% assistance). On steep climbs where I would use level 3( 100% assist) and level 4 (200% assist), I would average about 20wh per mile. A power meter is an accurate way of gauging efficiency.

JGaerland, I use the same wattmeter. They're quite repeatable on the same unit, but a 10% variation between units is typical. They're also fragile. I have about three just quit working. Oh well, I buy them on ebay for $11-12.

As I wrote earlier, if I ride at a slower pace, about 8WH/mile at 10-11 mph. It goes up with speed.

You have to compare motors with the same type of sensor as a cadence sensor provides constant power as you cycle but a torque sensor provides proportional power to how much effort the rider puts in with minimal power in many situations.. Some hub motor based ebikes with torque sensors can deliver many more miles than their cadence sensor based versions like the Suntour HESC system. Just thinking of the Carrera Suntour HESC based ebikes based in the UK. They have models with only around 300Wh batteries claiming up to 60 miles but for a standard cadence based sensor ebike with a similar hub motor I've often seen 20 miles given. The Suntour HESC system too its hub motor is rated around 60-65Nm which is typically higher than most cadence based hub motors. Perhaps because its a torque sensor they can allow a little more power for short periods of time. However they feel different. The cadence sensor ebikes feel more powerful as they are assisting constantly where as the torque sensor ebikes make it feel like you have bionic legs. It's the same effect on a hub motor as mid-drive motor as long as you have a torque sensor. The Suntour HESC system also feels much better at climbing hills compared to the low power city bike mid-drive motors from Bosch etc but of course doesn't match a full on e-mountain bike with a more powerful mid-drive motor which also has lower gearing to maximise that power those are significantly stronger at hill climbing.