As a child I collected photo encyclopedias of things like space, and they all universally said the major reason the N-1 failed was the multiplicity of engines a

As a child I collected photo encyclopedias of things like space, and they all universally said the major reason the N-1 failed was the multiplicity of engines and the complexity of controlling them, along with reliability. Likewise the OTRAG proposal (bundles of simple rockets) which was killed by politics, but remained essentially untested except for ground tests. I now begin to suspect it was a join over the pre-history of computing, and post-hoc reasoning behind the F-1. What worked? 5 big ones. What else might work? We don't know. the ones who tried lots of smaller ones failed. It does look like a highly reliable rocket cluster can work, if you have the smarts and can handle flame-outs. The V1 and some follow on had vanes in the thrust plume to steer. Crude, but also possibly reliable? They've been rejected for thrust vectoring but that means flexible joins and actuators for big flamey things with lots of energy and weak joint risks. Those flappy paddles which come out of the SpaceX booster look like something in the same vein: Simple, mechanical methods to get some airflow alignment and manage things without having to make it .. complicated.