Compiling RTL to run on an FPGA is way more complicated than compiling code to run on a CPU. Typically it has to meet timing, which requires detailed knowledge of logic placement. I'm not saying that's impossible, just that it's more complicated.
So yes, in that sense I'm talking out of my ass. But perhaps you can help enlighten me what it is that makes building FPGA firmware different from building MCU firmware.