Corner-Cube Mounts
We need at least one corner cube mounted on the primary mirror cell overhanging the primary mirror and looking down at the mirror. In this way, some of the light leaving the telescope on its way to the moon gets intercepted and sent back to the receiver for measurement of the pulse departure time. Ideally, we would only use a single corner cube to accomplish this, but if we had multiple corner cubes at staggered heights above the primary we would gain statistical power in an unbiased way (we could in effect get stronger departure fiducial signals without the many-photon early bias being a problem).
We need at least two corner cubes on opposite sides of the primary to verify beam divergence as it leaves the telescope. But we only need these two (or more) occasionally. So we could either:
Point 1. has the advantage of only having one corner cube to mount carefully and survey with respect to the telescope axes. Point 2. has the advantage of a strong fiducial, but the disadvantage of having many to mount carefully and survey (but maybe this is actually an error-reducing advantage). Another disadvantage to many rather than one is the obscuration/scattering/diffraction that may impact other observers.
Regardless, we may want to make the "permanent" mount a kinematically repeatable but easily removed mount, so-as to facilitate work on the telescope (people walking around the mirror cell). The absolute simplest scheme for the removable type would be a magnetic base attachment, possibly with a safety cord to prevent accidental falls onto the primary mirror.
The corner-cube mount itself may be commercially available, but we'd want to pay attention to stray light and scattering issues, especially for a permanent mount.
As for location, there are eight spots around the primary mirror where we could fit between mirror-cover petals. As for corner cube size, this needs to be worked out for sure, but I believe the optimal size will be in the neighborhood of 1 or 1.5-inch diameter.