With it happening every 5 minutes I would honestly focus on that portion of your issue, and track that down to see if it leads to the issues. A timed variable in a computer environment should be easy to find.
Also, you said something that caught my attention by switching to Windows Audio; it uses more resources. The only way that should happen is if the buffer is set differently than the previous audio setting. Unless the specific audio device has internal processors dedicated to decoding audio, your CPU still does the decoding.
You say you're using a Surface 4, I would research what built in software is running any automatic sequences. It honestly sounds like either something updating every 5 minutes and/ or something trying to save power.
What I would do. Download Revo uninstaller (free one). Check auto start programs and disable ones you don't need. Also, I'm not too familiar with a Surface pro, but if it's running Windows I bet a ton is working in the background.
For example in Windows 10 REVO I disable just about everything, except for Chrome update ahha. You can also remove services Windows won't allow you from the OS with that software, so be careful.