Yes.
Outside the context of this problem, I have even considered the very idea potential energy, in general is an illusion manifested by kinetic energy contained below a certain bound of observation. For this problem, I have seem to adopted the notion of potential energy.
I can now imagine the paths of, say two particles of air, moving in two hyperbolic paths, approaching at a minimum and diverging away (like the "hourglass" figure). The particles change direction, but not necessarily their speed. The result of the latter divergence is to expand the volume occupied by two particles, allowing force to be applied to a surface, which is itself moving in the forward direction (from the point of view of the center of momentum of "two particles+surface") while the other two particles whose sum drifts in this backward direction.
Then all I have to do is increase the number of such interactions to extrapolate how the kinetic energy can work on the sail in the opposite direction of the overall drift even though it is moving in the opposite direction.