You don't sound convinced. Okay, let's review. For galvanic corrosion to occur, it requires dissimilar metals to be touching each other in the presence of an electrolyte. The metals can be expressed in terms of being "noble" (ability to resist corrosion) and "non-noble" (tendency to corrode). This is called galvanic potential, and it can be expressed in terms of voltage differential across the two metals in the presence of an electrolyte. Take a look at the chart near the end of this page (anodic index):
Galvanic series
Looking at the relationship between plated/chromium (stainless) steel and plain carbon steels, the voltage differential can vary between 0.15V to 0.30V, which places it in the "acceptable outdoors" to "monitor/caution" range, depending on exact alloys used and conditions present.
For example, on my roof rack, I've used 316 and 304 stainless hardware to assemble it- this actually places the voltage differential (and therefore galvanic potential)
higher, in favor of the hardware. Using your example of salty sea air as an electrolyte (I also live near the ocean as well as visit mountains during ski season), that means that my roof should be corroding to pieces around the more noble hardware.
However, that isn't true- see the picture below for a picture I took today, with the hardware and roof looking the same as the day it was installed. The reason for this is because the metals are isolated from one another as much as possible (the sheetmetal on the truck is painted, the rack is powdercoated, the threads where metals might touch have thread sealant applied, etc.).
I have also made sure to seal the interface using weather-resistant rubber washers and butyl rubber sealant, which allows any water (electrolyte) to drain and evaporate away from the area as quickly as possible, in the same way that the factory roof rack does. This ensures that even with a galvanic potential present, it's isolated from the electrolyte (water).
This is important, because using silicone/UTV sealant (and especially when it's poorly applied or applied several times) may not provide the dynamic seal needed for heavy use outdoor applications. Moisture creeps in to the small crevices that are not sealed, and stagnates in the threaded holes in the sheetmetal. You don't need dissimilar metals to begin oxidation in this case- steel will rust on its own accord. Using your logic, I should have selected hardware with less galvanic potential, like plain uncoated steel. But that would simply accelerate normal oxidation and the hardware would rust due to exposure to the elements.
To sum up: yes, galvanic corrosion is real and must be accounted for in certain applications. However, isolating metals and preventing moisture buildup is even more important. If you understand and apply practical methodology when installing a roof rack, you will be fine- the same way Toyota is fine when they install the factory roof rack. Toyota builds and sells thousands of 4Runners every year, and very few have Swiss cheese roofs. If you poorly seal the roof rack mounting locations, no amount of galvanic similarity will save you.