I ran simulations and here's the raw data. We're looking at the velocity when fuel tanks hit empty rather than max altitude because it's more linear for percentage comparison. (Max altitude would be exponential since gravity's capacity to decelerate decreases the further away you get) Added weight for cones is not accounted for because we are looking at their capacity to increase efficiency and thus "be worth the extra weight". Structural pieces' weight is considered negligeable (-2.6% for 3 1x4, thus likely -0.9% for the 4 1x1)
Objects were launched at 100% thrust, 100% vertical.
Straight, No-Cone "Control" : 100%
View attachment 363
A-frame, Exposed-Gap : 90.9%
View attachment 364
A-Frame, Enclosed-Gap : 90.9‰
View attachment 365
A-Frame, No-Cone : 92.6%
View attachment 366
A-Frame, Coned, Structural for reference : 112.4%
View attachment 367
A-Frame, Coned : 114.2%
View attachment 368
A-Frame, Directinal Coned : 114.2%
View attachment 369
Straight, Coned : 122.8%
View attachment 370
These results are most unexpected. I'm surprised that Straight is more efficient than A-Frame ; and that a gap under the cone (still covering 1/3 surface area) is worse than no cones at all. I'd be tempted to say "bug?"