As far as marimo is concerned: this in fact was inspired by Pluto. It’s main advantage nowadays however is that notebooks can be deployed via wasm with zero installation efforts for the user, but this would be another thread.
I see the loading time issue as critical. Extrapolating the time between 1.6 and 1.10, we are now 9 months away from another LTS release (1.13 ?) which would have e.g. public, and [sources], [workspaces] in the project files with the potential to significantly improve the sustainability of the package ecosystem. As far as I remember, SciML was quite quick to require 1.10 as minimal Julia version, and the additions I mentioned are good reasons to perform a similar quick transition for the next lts. Sticking to 1.10 for package loading time issues then would hold back the whole package ecosystem.