Overflow of JES2 Resources beause of typing error in a batch application or in using the FTP JES2 interface

In a JES2 environment it is easily possible to quickly overflow every single type of JES2 resources, which are JQE's, JOE's, BERT's and Spool Space.
A simple typing error in a batch application or in using the FTP JES2 interface may lead to triggering millions of jobs resulting in JES2 resource exhaustion in some minutes.
Coming into such a situation prevents creating any new TSO sessions, submitting batch jobs or starting Started Tasks. If nobody is currently logged on to the system it will take a very long time until the root cause of the problem gets detected. And it will even take some more time until the suitable operator commands are found to clear the situation.
In a production system such a situation may lead to disastrous effects to the business.
The new JES2 emergency subsystem (“Privileged Separation”) that came with z/OS V2.3 keeps aside a separate pool of resources for authorized users, but until a user is available to fix the problems it may take a long time.
Obviously there is no method available in JES2 to cope with such kinds of situations in an efficient way which means to eliminate the problem at the same time it arises.

We need an internal component in JES2 (or z/OS) that is able to detect such kind of situations and react quickly to those.
Once the situation is detected, there should be an internal mechanism within the system to halt the flow of job submits. No further jobs should be entered into the JES2 subsystem and constrain JES2 resources from then.
There should be a JES2 or operating system parameter by which this mechanism may be enabled or disabled by choice.