BTW, Martin, a central motor installation isn't all that much more expensive than a front motor. It's the bigger battery you need when you go from a lower power front motor to a beefier mid-motor that raises the comparable cost appreciably. Also, the front motor that I had, despite its reputation as the Little Hill Climber, seemed more lightweight than the central motor, in every aspect but especially the control electronics, which were clearly not an in-house production in the case of the front motor. It might be that the front motor was built light to earn its title on the hills. Don't get me wrong. The front Bafang motor lasted lonhrt than any of my Nexus boxes, and the central Bafang motor seems likely to see me out, so I don't have any complaint about value received; I'll buy Bafang again, if necessary. But, if you're planning to make a permanent installation in the first instance, I suggest the central motor would be more satisfactory in the long term. The reason I made a sacrificial front installation was because there was very little known about these motors at the time (except for power-crazy Americans building electric motorbikes), so I was educating myself, but that no longer applies. There's a lot of experience of electrical motors on bicycles on this forum and elsewhere.

Cost isn't much of an issue. OZO is a French supplier who seems to know their subject, they assemble their own batteries using Panasonic cells and supply batteries for several different uses (electric bikes, small boats, storage batteries to allow stockage and restitution of electricity from solar panels). They also have a stock of spare parts to rebuild and maintain Bafang motors.

From OZO, a complete 250W 40nm front wheel motor kit costs nearly as much as a BBS01 Bafang 250W 80nm mid-motor kit. The mid-motor kit is heavier, with double the rated torque. The front motor kit is simpler, so easier to install and probably also to maintain, and about 2.5 kg lighter. Weight on an electric bike isn't really an issue for me. But I would like to try and convince my wife to try an electric bike, and she finds a light bike easier to handle.

I have already ridden about 180 kms on the 2 second-hand mid-motor electric bikes I sourced for the nature reserve. These have the Bafang M200 integrated motor, which is similar to the BBS01. Performance of these bikes is quite sufficient for what I think I will need from an electric bike.

But I still intend to try a hire bike with a front-wheel motor to see what the advantages and disadvantages of each system are while riding.

A bit late in the day, but I thought I'd best write a follow up.

I opted to change the combined brake/shifter levers on the flat (butterfly) bars for cantilever specific ones, a Shimano Deore LX M567 3x8 Speed Shifter and Brake Lever Set and update the cantilevers to Shimano Altus BR-CT91.  The bike brakes and shifts beautifully now though at some additional cost, but I'm enjoying it.

Thanks for all the advice and help

Thanks for reminding me, Martin:

I should've mentioned that the constraint behind my choices of chainring tooth counts is those that take Chaingliders, which are simply too useful to give up for some notionally "perfect" gear ratio; there is always another way to skin a cat.

And equally, I aim at cruising at my comfortable cadence of 40rpm on the level in gear 11, which is a 1:1 ratio, saving overdrive gears 12, 13 and 14 for fast downhills. Not that I'm on the flat all that often, but if you set up a Rohloff so that 11 is your most used gear, a lot of other possibilities fall into place -- for instance, after I fitted an electric motor, I could avoid the Rohloff's lower 7 gears by using the motor's software "gears" if I wanted to, though I don't.

You may ask why I worked up a complicated table for myself with a cadence of 60rpm when I'm comfortable at 40rpm. The answer is that 60rpm multiplied by 60 minutes in an hour has a red notch on a pilot's slide rule, and I'm comfortable working mental arithmetic on a base of 3600. For instance, 40rpm is 2/3 of 60rpm, so I eyeball a number, say a speed, on a different table and can instantly see how realistic it is; or I could argue that 2/3 of 3600 is 2400, which is pretty close to a quarter of 10,000, actually 9600, close enough for a first approximation or a check that you're working in the right direction; you can see that it makes mentally shifting the decimal point a lot easier. Martin's cadence of 90rpm is one and a half times 60rpm, so the result should run in the other direction, an easy division by two and then adding the product back to the base, result same as multiplying by 1.5. To be clear, I don't have the eyes any more for the small print on slide rules or the rotating slide rule bezel around a few of my watches, but I don't really need to see them; from long use I effortlessly picture them in my mind's eye. I fondled my pocket Pickett a bit and then slid it back into its leather sheath.

***
BTW, Martin, a central motor installation isn't all that much more expensive than a front motor. It's the bigger battery you need when you go from a lower power front motor to a beefier mid-motor that raises the comparable cost appreciably. Also, the front motor that I had, despite its reputation as the Little Hill Climber, seemed more lightweight than the central motor, in every aspect but especially the control electronics, which were clearly not an in-house production in the case of the front motor. It might be that the front motor was built light to earn its title on the hills. Don't get me wrong. The front Bafang motor lasted longer than any of my Nexus boxes, and the central Bafang motor seems likely to see me out, so I don't have any complaint about value received; I'll buy Bafang again, if necessary. But, if you're planning to make a permanent installation in the first instance, I suggest the central motor would be more satisfactory in the long term. The reason I made a sacrificial front installation was because there was very little known about these motors at the time (except for power-crazy Americans building electric motorbikes), so I was educating myself, but that no longer applies. There's a lot of experience of electrical motors on bicycles on this forum and elsewhere.

If a cyclist has to push his bike up a hill because he has run out of puff, it is because his lungs can't supply the oxygen his muscles demand and/or because his heart can't circulate his blood fast enough to deliver the oxygen to his muscles. In all cases it is because something in his transmission was badly chosen, usually too optimistic.

All of this relates through the cyclist's maximum heart rate to his cadence. His theoretical maximum heart rate as a first approximation is roughly 220 minus his age, but that notional maximum is a number only professional athletes under medical supervision should approach. We generally speak in percentages because everyone's maximum heart rate is different. Ninety percent is possible for young men but you don't want to be in the 80 percents because there lactic acids are formed which will undo your good work. So for most of us, and especially for middle-aged cyclists, 75 per cent of age-related maximum is a reasonable heart-rate if you wear a heart monitor and pay attention to its output, otherwise lower still. If you're a careless rider, pick something even lower.

After heart surgery and physiotherapy, I found that the easiest way to relate this is through the cadence I could maintain on the flat for several hours. (I'm never on the bike for more than a couple of hours without a break.) Heavy tourers and cyclists who want to cross a small country in a day could pick that long day's best maintainable average cadence.

But all this knowledge will avail the cyclist nothing if the tooth counts on his sprocket or his chainring are optimistically chosen. The cyclist must start with a rationally and realistically chosen consistently achievable cadence and match the tooth counts on the gears to that cadence.

On this page-- 
http://coolmainpress.com/BICYCLINGHebieChainglider.html
you will find tables for
--60x622mm tyre at a cadence of 60rpm
--26x1.6in tyre at a cadence of 80rpm
--26x1.6in tyre at a cadence of 120rpm
--26inx50mm Schwalbe Dureme at a cadence of 65rpm

You can interpolate between the tables with only two facts: the circumference (standing in for the rolling length) of your chosen tyre and your cadence, if different from those used in the tables listed above. You don't want gears below your ability to balance the bike, and you don't want gears that force you to pedal faster than your reasonable cadence maximum. Coasting on the downhills is a wonderful opportunity to enjoy the surroundings.

After all that work, I found that on my preferred 622mm rims clad with 60mm Big Apples at my preferred cadence of 40rpm I liked the ex-factory default of 16T sprocket with a 38T chainring, super for hauling heavy easels and paint in my all-hills countryside in West Cork. Since fitting a torquey electric centre motor, I've moved up to a 42T chainring.

I recently sought some advice re gearing down. The Clisham, my Hebridean nemesis means my 45 x 19 is a bit too much for me ( loaded Nomad etc.)
I’ve had the suggestion that dropping down to a 39T chainring and keeping the 19T sprocket will give me a couple of lower gears. I’ll give that a go and hopefully conquer The Clisham next time I’m there.

I haven’t looked into cadence etc.