I think i might have worked out the issue at least whats been grinding the server to a halt the past few days.

This morning i came in early before there were many users on and logged in and saw Windows task manager reporting still avout 30-40% CPU constant. I was like what the.... Than i saw that when the server was rebooted on the week end just past something had gone wrong with the data collection jobs and one was stuck in a cancelling state since basically SQL was restart. When i force cancelled this just all CPU cores just dropped to 0% util, Been monitoring the past 3 hours and not seen any spikes longer than a few seconds above 30% which is more normal for our environment. Certainly not the 70-90% mark i had yesterday.

I still want to make those few changes to cost of parra and maxdop what is a safe starting point for cost? most seem to say 50?