There are some servers that don't get started in the CONFIG instances - there's a 'jobmanager service'. I think that will still run even if you turn those settings to 0. So there would be some overhead I imagine.
What's the use case for this? why do you need to make these changes like this?
Well, no specific day to day use case, but:
1) It's easier to install: You do not need to disable the current instance and create a new one. (minor advantage)
2) It's easy to play with scalability. If for some reason the Job appserves are fully loaded one could choose to give the console servers some work as well. During weekdays we have a lot of console users preparing releases and a small number of installations, but in the weekend it shifts the other way around. A limited amount of users is working in the weekend to install the packages, but they will generate a lot of jobs.
But you're going to have to restart the appserver to have those changes take effect.
If you are worried about capacity, I'd create the additional config and job instances and then stop/start them as needed. I'm assuming the config instances are behind a load balancer. Not sure if you can do that from the command line or not. in this way you are not restarting the whole appserver, just the instances you don't need.
Or I'd just get more hardware to handle the peak loads...