CLI pipeline inefficiency

I was looking into an issue with the ```azure-cli``` repository pipelines and I noticed as part of my investigation that the pipeline appears to use a lot of separate jobs. It seems like there are stages that all run on the same vmImage which could execute sequentially as tasks within the same job.

This would result in less resource wastage (VMs needlessly being drained from the shared pool) and probably better performance (less overhead spinning up tools on a cold VM).