I've learned a lot from experimenting with different designs of fan duct. My previous experiment, the 360 degree fan duct, didn't work out as well as I had hoped. However, this fan duct works fairly well. I learned how to set up flow simulations and used that to help me tune the performance of the fan duct. I combined this flow simulation testing with real world testing to determine if it meets my design goal. 

The goal with this design was to send as much air towards the tip of the nozzle as possible. The intention of this was to improve cooling rate on smaller prints. In my testing, I found the limiting factor is mostly print settings and as long as the fan shroud does its job, the print quality can't be improved beyond what your slicer settings allow. That being said, this fan performed better than stock in most of my testing. Bridging tests resulted in (marginally) less dangling and benchies printed mostly flawlessly for both fan shrouds, but were only slightly more stable on the front overhang with my fan shroud. 

Overall, my fan shroud did outperform the stock fan shroud in my testing, but not by an earth-shattering amount. Feel free to test it yourselves and share your findings with me.

The part can be freely modified by following the link below:
https://cad.onshape.com/documents/a7e64a740cdc3aa4d71dcca2/w/8a8a5974a2c91f518ba4a4dd/e/72afdd29311311e2e6b82430