It might happen that script 1 in part X crashes all the time.
Script 2 will then never be executed.

(this has hjappened btw in the past with ez's own notification cronjob blocking execution of workflows)

If we separate execution of every single script in its own process, this can be avoided.

It also helps if script 1 leaks memory.

In the extreme case, we could even launch all scripts in parallel.

ezfind already does this...