Ok this is a long story that I've investigated and experimented with for years.
To cut to my opinion of the cause, the forks clunk when they top out for two reasons, they have a topping out bush rather than a spring which is very unusual and the rebound damping is insufficient particularly over the last inch of travel when the rebound damping should increase substantially to provide a bump stop effect but doesn't or didn't on mine.
Here’s how the forks work.
Compression Damping
During fork compression oil is displaced from the lower fork leg into the stanchion. It passes through the compression damping holes in the bottom of the damper rod and also through the annular gap between the damper rod and valve body lifting the valve washer up against the valve plate. The degree of compression damping is determined by the size of the compression damping holes and the oil viscosity
Rebound Damping
As the forks extend the damper piston forces oil back into the lower fork leg. The valve washer drops and seals against the valve body and the oil can now only pass through a small 3mm hole near the top of the piston damper rod. This is the rebound orifice. The area of this hole is smaller than the area of the compression damping holes by a factor of 3 resulting in approx 3 times stiffer rebound damping relative to compression. That figure is typical as rebound damping is always substantially stronger than compression damping. During the last 25mm or so of extension this hole drops below the valve washer so there is nowhere for the oil to go except by leakage past the washer and to a much lesser extent past the piston rings. This is meant to be the rebound hydraulic bump stop and its efficiency will depend on the leakages mentioned. Variability here may explain why some owners have problems with topping out clunk and others do not. In my case the area of leakage past the valve washer was twice as much as through the rebound hole proper so there was very little effective hydraulic bump stop effect. In making my own valve washer I reduce the washer to piston rod clearance to cut the leakage in half and this eventually eliminated the topping out noise when I got the clearance right. Why didn't BMW get this right in the first place - I think it's because you can go too far in reducing leakage and the forks will tend to hydraulic lock on rebound.
When the thicker chamfered valve washer was introduced the washer to piston rod clearance was reduced which reduced leakage and therefore improved the rebound hydraulic bump stop.
What I strongly recommend you do is remove the springs leaving the oil in, then stroke each leg by hand to feel the damping. You should feel compression damping hardly at all but rebound damping should feel stronger and there should be a definite increase over the last inch of travel If you don't feel that increase then it's going to clunk for sure. The cure is to reduce the leakage past the valve washer by closer tolerances. BMW modified the valve body and washer several times in an attempt to improve the forks The first iteration was a solid valve body and steel washer which is what I had, then they changed the washer for a thicker plastic washer with tighter clearances and less travel, the recess in the valve body was made deeper to match. Another source of noise is movement of the valve body in the stanchion, the valve body was initially loose then they shimmed it to prevent it moving, then used a sort of sprung washer and eventually a shorter sprung valve body was used. All these things were an attempt to stop the valve body rattling but the sprung version may also help with the clunk too.
You are absolutely right to suggest that stronger springs will make the problem worse by overcoming the rebound damping even more.
In my opinion an R65 is the last of the airhead models to need stronger springs unless the rider is well above average weight.
