The MLAS flight plan, as it eventually settled out, is shown to the left. I thought I could capture a good part of that with µMLAS ... certainly launch, separation of the skirts, ejection of the CM, and recovery. If I stuck to 1:20 scale and added the skirts, the resulting model would be significantly heavier than µMRA. So I took a leap (for me) and opted for 29mm motors. The G80 was the limit without the expense and hassle of a BATFE Low Explosive User's Permit (LEUP), but at this point that seemed more than adequate. I wasn't ambitious enough to attempt mounting four 29mm motors in the boost skirt, canted through the centerline MLAS-style, but thought I could cluster three easily enough in a 3" motor mount the conventional way and still leave room in the Coast Skirt (CS) for electronics and a parachute compartment. The Boost Skirt would drag separate at burnout and be recovered on its own chute. A timer in the CS would deploy an aft-mounted chute, MLAS style, to separate that section. I'd forgo the turnaround maneuver and use a second timer in the CM to separate it from the Forward Fairing (FF) and recover both those sections on their own chutes.
I crossed the line into high-power rocketry when I opted to cluster those motors, of course. But a smallish 8-fin-nose-cone model seemed a little boring and the complex large-scale option was within reach ... put another way, it seemed like a good idea at the time. My early RockSim runs said I could expect 1000 feet on a cluster of 3 G80's, but my estimate of the weight wasn't much better than an educated guess. In retrospect, I should have controlled weight more carefully during the build, but I wasn't being retrospective at the time and thought I had lots of margin. There was the issue of stability to consider too, but I could address that with ballast ... after all, I had lots of margin.
As I said, it seemed like a good idea at the time.