"What about driving wheels?"  I am standing in the bar at a National Model Railroad Association convention, an ice-cold Coke in my hand.  The grey-bearded craftsman sharing the break with me has been saving this question, I am certain.  Sure he nodded approvingly as I explained the virtues of scale track; sure he agreed with me about the vices of steam roller wheels.  I thought I had a convert until he played this trump card.  I reward him by looking sufficiently uncomfortable, and perhaps blushing a bit.
The hard truth is there is no simple answer. There seems to be enough difficulty in finding drivers for standard HO scale, let alone for a niche such as this.  To make matters worse, Proto:87 modellers are the type of people who are likely to be particularly finicky when it comes to drivers.  Never satisfied with just the right diameter and crank throw -- enough to make it operate correctly with the rest of the locomotive -- we're likely to want to see the right number of spokes, relationship between spoke and crank, and yes the the correct pattern of spoke, bevel and probably casting marks as well.
Needless to say, it is going to be a while before Alan Gibson, the lone supplier of Proto:87 steam locomotive drivers, gets around to making the varieties we require.  So if you're waiting for the ready-made answer, prepare to wait a while, or consider getting excited about diesels, which you can already model with off-the-shelf components.  It's no use getting upset about this.  Alan has been a better supporter for us than we have been for him.  Each driver he makes available represents a considerable cost, and so far all the drivers that have left his premises could fit in your cupped hand.
It's an unfortunate Catch-22.  On the one hand, steam modellers won't switch to Proto:87 unless sufficient drivers are available to re-wheel their stables.  On the other, manufacturers won't make a wide selection of drivers unless there are enough customers so they can recoup their investment.  The size of that investment is the key; injection moulded plastic centres with machined rims are an

expensive form of wheel if you only need a handful.
A number of manufacturing techniques would surely be more efficient for the size of run we need.  In some other scales modellers have offered to turn commercial wheels to a new profile for a fee; many members are already capable of offering such a service.  Someone with a little more ability could offer pre-turned tyres with resin centres.  Finally, machined lost wax castings do not require many more tools or abilities than many of us already possess. 
While the unit cost of any of these techniques is quite high, there is a low initial investment.  That means there is little risk for a potential manufacturer.  On the other hand, it also means the price of these drivers would likely be quite high.  But then, no-one needs more than a handful of drivers anyway.  The conditions are ripe for a cottage industry.
Cottage industries provide us with the stepping stone toward more viable Proto:87 steam modelling.  Indeed, if you consider just about every other breakthrough in railroad modelling, it has come not from the established manufacturers but from individual modellers who have produced something for themselves and then offered it to others.  Some of those cottage industries have grown up into the Kadees, Athearns and indeed Alan Gibsons of today.  However, if we want Proto:87 drivers, then a Proto:87 modeller really needs to step up to the mark and produce the initial offerings.
I am sure there are several of you who have the ability and interest to accept this challenge, and I look forward to seeing early samples of your work.  For now though, in this issue, Roger Miener and I explore the possibility of borrowing some hand-me-down wheels from our big sister, P4.  This approach introduces us to a dizzying selection of over four hundred wheels.  Using bastardized P4 wheels, of course, is not going to yield completely accurate models, but it does facilitate the beginnings of steam modelling..