What I'm seeing is way too much "compression", which technically is the wrong term. What we call "compression" is actually bit-rate reduction (or bandwidth reduction), a means to reduce to data, per second, required to render video. All video codecs, going back to the first QuickTime of the early 1990s, attempt to reduce the amount of data by encoding only what changes, reducing spatial and temporal redundancy, frame to frame. It's a bit more complex than that, but that's the simple explanation. Elements that don't change frame to frame are not re-encoded, they're simply repeated. What burns up bandwidth in video are two aspecs: detail and motion. Both result in very little redundancy, so reducing the data has a negative visual impact, loss of clarity, blocking, etc. What you have in a shot moving over water is pretty much everything changes in each frame, so if the codec target is very low bit-rates, and every pixel is different in every frame, the codec has to throw out lots of detail and non-repeating information just to hit the target low bandwidth rate.
Yours is entirely a code "compression" issue. Increasing the frame rate won't really help because that increases the base bit-rate, and the codec will have an even harder time reducing it to the target rate (assuming that is baked in and you have no way to set the end bit-rate). However, reducing the shutter speed with an ND filter will have a positive effect because the result will be motion/speed blur, which reduces detail, and is actually easier to encode. When a high shutter speed is used, motion blur is reduced or eliminated, meaning each frame is highly unique, and what pixels would not not encode then? They're all different, every frame. A side effect if increasing the frame rate to 60fps or 120fps is the shutter speed MUST be higher. Again, going the wrong way if the end result must be run through a lossy codec.
If Premier Rush won't let you tweak the output codec's target bit rate, then don't output that way. Export the video a in a way that results in bigger files, higher bandwidth, even full 4K, and just let YouTube do it's thing, which usually isn't that bad. In other words, don't output directly out of Rush to YT, go to a high quality file first, then upload it. Your video is getting hit with bandwidth reduction twice, once by YT and once by Rush.
The only other thing to consider is shooting using H.265, which is a far more efficient codec to begin with. It results in higher quality video for the same bandwidth or file size, but with less codec artifact.
So, h.265, 4K, 30fps, and lower the shutter speed with an ND filter. Get it down to match the 180 degree rule (30fps would be 1/60th shutter). Then edit in 4K project, export in 4K with as gentle a codec as you can, let YouTube be the main compressor.
