Do I really need to run compensation controls everyday?

Yes. Even though the instruments are QC'd on a daily basis to measure stability and insure we can provide researchers with platforms for reproducible data, there are always small differences in the instruments and more so in your experimental set-up each day. Under proper set up you will notice small changes in instrument settings from day to day, so small that you might conclude that Set-up Controls aren't really that important everyday. This is faulty logic, when you see differences in your Control Sample compared with your Experimental Samples, you'll want to be sure it's really due to your design, and not due to some errant fluctuation or improper set-up.