Lou,
I don't understand how pressing a roller into a tire will place stress on a spoke? I have never bent nor broke a spoke and have many years, unknow mileage but pretty high, so I feel I have good RL experience. This system "floats" on the tire as it flex's from eyelets.
Happy Time, I suggested that you had snapped to a decesion as you are describing faults that don't occur with the kit style I use because of its design style. Not so much as defensive as frustrated in that folks here always have an opinion and tell me what is wrong with this kit style.... and it doesn't have those issues. Its not like a Staton at all except in curosary looks, operationaly its night and day different.
Now that I have been here for a couple months I think I understand how the Staton works and I will post some new pics of my bike showing why its different and "kinder" to the bike parts. A differnet thread really explained the parts on the other,[Staton] kits and I realise now that they seem to operate more like the old "Island Hoppers" did. If that is correct I now understand some of your issues with them.
It's not going to put any unbalanced side-to-side single-point stress on any spokes, like clamping a pulley or sprocket to the spokes will.
However, it will put a radial downward force on the wheel, at the point where your friction wheel is pressing into the rubber. This force acts to distort the wheel. Normally, the weight of the bike/rider provides the only radial strain on the spokes. This force is upward from the ground, and the wheel will deflect slightly, and become very slightly egg, or pear shaped. The bottom spokes will become slightly less stressed, and the side and top spokes will be under slightly more strain. If you're pushing down against the wheel as well from the top, you'll end up with a wheel that is slightly oval shaped, which will ever-so-slightly stretch the leading and trailing spokes, which face the front and rear of the bike. Again, because the additional strain is shared by the rest of the spokes which are not 'load bearing' at the top and bottom of the wheel. As the tire rolls, the strain shifts smoothly, from spoke to spoke. With a non-friction drive bike, as a given spoke rotates, it reaches minimum strain at the very top of its rotation. It reaches maximum strain when pointing straight ahead and straight back, (when horizontal to the ground) - ref the attached strain plot. By adding downward force at the top of the wheel, this will tend to make the 'valley' on that plot a little shallower, and make the 'hills' on either side (at +/- 90 degrees) a little taller. This can be explained by looking at the second image, a quick, sketch of the wheel/spokes, with the deflections GREATLY exagerated... Because a force acting on a circle tends to deflect the circle into an oval, the points at 90 degrees from the force will have the greatest deflection, and therefore, the spokes will have the greatest strain at 90 degrees away from the force.
If you could move the friction wheel to the front or the back, it would tend to reduce the strain caused by load bearing. If it's at the top (or bottom) it tends to add to the strain.
In addition, the action of the motor will add torque to the wheel at the top. That torque is transferred to the ground, through the wheel, and the spokes. Since the torque is evenly distributed amongst all the spokes (assuming even tension on all the spokes) there really isn't any more net torque induced strain on the spokes than if you were providing the torque by peddling.