2) Depends entirely on the capacity of your platform... disk, memory, network... Nobody here has visibility to your platform. You'll have to work with your system guys to optimize workloads and throughput.
If you start 10 jobs at exactly the same time then one or more can fail. If you start them a few seconds one after another then you can run more jobs. The start up time is very intense and consumes a lot of system resources. Some systems 10 jobs at once is no problem but idea is the same. Start up is worse or more intense than when the jobs is running.