I did, but I went a somewhat unconventional route. I designed and 3D printed a replacement fan shroud that holds three 140mm fans and populated it with Noctua NF-A14 PPC-3000 PWM fans. The setup seemed to work well. The 8x 15K SAS HDDs I had installed hovered around 35C-38C with low to moderate use, which remained GREEN on UnRAIDs dashboard temperature-wise. The SSDs that were installed were inconsequential. The room its in is around 23C. Unfortunately, the server is torn down / offline at the moment for an unrelated rebuild, so I'm unable to pull CPU temps, etc.
Noise-wise they remained audible at all times, but it was a low background hum that was almost unnoticeable. In fact, I didn't realize how used to them I was until I shut the server off ... the office is creepy quiet, and the white noise from the fans really helped drown out neighbors, etc.
The NF-A8s are far too wimpy, unfortunately. 😞
What I've found interesting is the different fans that different server/case manufacturers build into their chassis. It makes me curious where their data comes from, what kind of tests they perform, what they design for, etc. For example, a 2U Supermicro box that I'm working on right now has 4x 80mm fans up front that are 11,000 RPM, 116 CFM, 54.3 mmH2O, 62.5 dBA monsters - for a 2U box with 12 drives. Meanwhile, the 4U UnRAID box I built with the same specs (same CPUs, same RAM, etc.) came with four 80mm fans with the specs in my original post.
EDIT: Both CPUs were actively cooled in my RM430 as well, which helped the Noctua fans pass muster.