A report summarizing this thread.
First, I would like to thank everyone who reached out to help. I appreciate the time and effort. Helping each other is what keeps our community here strong.
The basic problem was that while running on an inclined surface, the avi would fall through the surface. This problem occurred not at the ends, where one would suspect alignment issues, but in the middle of the surface. The problem occurred most frequently not when climbing or descending, but when running at a constant elevation. The problem was not isolated to one spot, but occurred at several locations. The problem was particularly pronounced when turning quickly.
For more background, the problem occurred whether the collidable surfaces were flat planes, court floors or surface waters, whether the surfaces were scaled to cover large areas or small, and whether or not there were overlapping surfaces. The area was carefully inspected, and there were no extraneous props with collidability that could have been causing the issue. The base elevation was 50000 in the y direction, but the issue also occurred at tests done at other elevations as well.
As a result of my research, I have concluded that the problem in inherent in sloped surfaces in RLC. The problem has never surfaced in my previous work, because the vast majority of slopes were stairs or narrow ramps, where one would naturally only move up and down. In this case, the sloped surface was a large inclined beach, where the most natural motion is to run along the beach perpendicular to the waterline at a constant elevation, rather that up and down.
The only solution available at this time is to redesign the scene. The beach itself is approximately 7000 clicks wide, covering 400 clicks of fall above the water line and 400 clicks below the waterline. Option one is to add rocks, piers, and such jutting out into the water with collidable surfaces underneath at the main elevation so that on cannot walk horizontally on the beach, but restricting access to the ramp to those areas where walking up and down will naturally occur. The second option is so shorten the inclined surface so that there is considerably less fall. The flat beach would be extended and the waterline raised. This will shorten the width of the inclined section, making it less likely that one will walk horizontally on it for long periods.
Should a reader find these conclusions incorrect, we would welcome the opportunity to improve our understanding.
Again, thanks to all who read this and offered their time and talents. Peace.