s_all has two ways to run commands - via the list of non-parallelisable scripts, and via the COMMANDS list. The COMMANDS list doesn't seem to reliably print the name of a command that fails, for example:
####################### MESSAGE ############################ s_all run of: "" resulted in: dist/type_to_str.py:36
What actually failed here is s_longlines because type_to_str.py has a long line in it, but that's not at all clear from the output. It should be printing:
####################### MESSAGE ############################
s_all run of: "s_longlines" resulted in:
dist/type_to_str.py:36
This might also be a problem with the non-parallelisable list too - checking this is part of the scope for this ticket.