Having finally made a rocket that's both powerful and flyable enough to take a shot at a return moon trip, I decided to give it a try. This is the result:
Yay! Considering this was my first go, everything went without a hitch. I learned a great deal about planning a landing approach for vertical landing. What worked for me was, I established an elliptical orbit around the Mun, with periapsis on the far side (relative to Kerbin) and at some 3000m height. Apoapsis was "aimed" towards Kerbin and at some 70,000m. This gave me a nice elongated orbital ellipse, so once I reached apoapsis I began my braking burn. This maneuver "narrows" the ellipse the marks the orbit, as well as shortening it - I made it so that the descent trajectory crosses the Munar surface at almost a right angle. Once at 20,000m height, I just oriented myself so the artificial horizon shows that my nose is pointing straight "up", used SAS to lock that down, used the main descent engines to slow my descent and RCS translation controls to kill any horizontal velocity. The reason I wanted apoapsis aimed at Kerbin side was, that side was also pointed at the Sun, and it helps seeing what you're doing when descending. Seeing your shadow is helpful, too
My descent stage actually allows for quite a big margin of error, there's plenty of surplus fuel, once I get the hang of landings I intend to bring less fuel and put a service module with some solar panels and a science package on the moon instead. Then detach and use the return stage for the trip home. This part is least problematic, as Kerbin's gravity does most of the work... and I find aerobraking a nice cheap way of reducing the apoapsis on my return trip. Quite like that trick. Even managed to land on water, smack in the middle of a nice bay, close to land. That would've been practical