One experiment i never got around to would be to make an adjustable motor separation assembly. This way you could separate the vertical motor distances as you read the output. I think this single handedly would have the biggest effect on performance and would need to be fine tuned for every setup. As everything in life has a resonating frequency you would need to dial it in for each individual setup. Somehow or other when NASA develops a coax it is more efficient. So clearly there is a way to make that lower prop "bite" into the dirty flow of air coming from above in a way that somehow benefits overall efficiency. I think making a cowling is a must as well to get max efficiency. maybe someone will one day come up with a way to economically manufacture these as they will not only protect things from hitting props but greatly add flight time. But what do I know. It has always been an interesting aspect of multirotors.