I've been reading up on RW, and am still confounded why RW should be more relevant than SW in the context of the double pendulum model and the notion of racquet lagging relative to an accelerating arm. Consider the following thought exercise. Does the RW of a hammer change if you swing it by the head versus the handle? It doesn't, because RW is calculated by integrating the mass distribution along the long axis, and with the balance point as the origin; it is not affected by whether you grip it at the top or bottom of the stick. And to follow, what do you think will be more laggy to swing, gripping a hammer by the handle or the head? It should be obvious that gripping a hammer the normal way will be more difficult to swing because most of the inertia is concentrated at the tip, creating lag. So here we have the problem that even though the RW of a hammer doesn't change regardless of how you hold it, the ease of swinging the tool is a drastically different experience if you grip it by the head. But do you know what does change in the hammer analogy? Yup, you guessed it; it's SW. The SW of a hammer when gripped normally is WAY, WAY higher than the SW held by the head. Just as Rod Cross concluded in his published article on the double pendulum problem, RHS is primarily sensitive to SW.
Conclusion and hot take: RW matters, but not in the way suggested by many in this thread. RW is relevant to recoil stability, and hence recoil weight. It is a measure of the resistance of a racquet to rotate around its balance point when a force is applied. When discussing swing mechanics, the SW is still the primary spec of relevance.