Anyone else finding the new Prime Tower behavior in V2.4.1 weird or is it just me?

I've been on 2.4.1 for about a week now with two four-color jobs under my belt, both around nine hours each, and the tower held up fine on both. No skips, no wobble, and the wipe wall stayed attached the whole way thru. I actually prefer the side approach over the old drop-down, my layer changes near the top of the tower used to leave a little blob where the nozzle smacked down, and that's gone now. Tool changes did feel a touch slower on the first print, but comparing the same model sliced on 2.3.0 versus this one, the total time difference was only a few minutes over eight hours, so I think the tower split eats most of that loss back.

I gave the new tool-change path two days before rolling back to 2.3.x on my Ray with the dual extrusion module. The side swing just didn't suit how I print. I run a lot of signage with dozens of swaps per job, and even a couple seconds added per change stacks up fast. My rough check of the gcode timing showed the new path costing noticeably more per swap, and a job that used to take me about six hours pushed closer to seven. That's real money in electricity and wear for me since I print for a small craft fair side hustle. On the paint penetration question, I only turn it on for parts with big flat top surfaces, like my coaster designs where bleeding shows badly on dark filament. For small brackets and clips I leave it off and just bump the flush volume a bit instead. Less waste, barely any difference in the result at that size. As for the tree support stuff, I've left the new parameters at default so far and honestly haven't noticed a change either way on the few support-heavy parts I've run. I'd rather people test on something expendable before trusting it on a long print. My two cents, some of these changes feel tuned for people doing big single-model multi-color art, not small functional parts.

I'm still figuring all this out, only got my machine a couple months back. The bleeding thing is my main headache too, I printed a name sign last week and the dark letters showed ghost outlines through the light front face. So that extra wall of a diffrent color underneath sounds like exactly the fix for me. Question though, does that setting only kick in where two colors actually touch each other, or does it add the extra layers everywhere even on single color spots?

On the tree support side, I ran a small bracket test yesterday with the minimum contact area left at default and the supports practically fused themselves to the top surface, took me longer to pick them off than the print took to finish, so anyone testing that parameter should probly start by lowering it rather than raising it. What I still have not worked out is how the tower splitting behaves mid-job, because my tower width changed on its own partway through when the adhesiveness numbers differed between spools, and I am not sure whether that recalculation is per layer or per filament. Has anyone watched a job closely enought to tell?

Update: took the tip about the tree support contact area and dropped it well below default on a small test print. Supports popped off clean this time, so that solved one headache. Still on the fence about the side swing though. For the folks running long jobs with lots of swaps, did you find a setting that trims the change time back down?