Let me clear up a common misconception. You are wasting time by stacking on stacks. Have a longer tray and that’s it. Stacking on stacks is an idea floated time and time again but it has yet to work out for any of the lift tray bots, me included. You’d need a boatload of bands to lift enough cubes to make scoring on top of a pre-existing stack worth it. You’re more than likely going to topple the stack. All of this is true of course only if your robot is able to remain steady and not itself topple from all that weight going up.
To be clear I’m not bashing you. Don’t feel bad for having the idea, it’s just that the reality is that stacking atop preexisting stacks is not feasible in my experiences with my lift tray. It’s possible, though. I’ve been able to stack 4 on 4 with a primitive version of my current robot. It’s feasible to score an 8, lift 3 cubes up, and then intake that 8h cube and score the stack.
Tray lifts main power is their tower swapping ability. Rollers are great at gripping those cubes a weird angles and the tray allows you to quickly swap a cube out in seconds, regardless if it’s fully nested or if it’s at a weird angle.
E: to add onto the design challenges, many robots don’t have a configuration where cubes can shift down when the tray is fully vertical, at least in my region. This allows the stack to be nice and straight on release, which is vital if you plan to put more cubes up. This is a challenge on a lift because it’s a balancing act between maximizing compression and compliancy.