You can see the generated code. When you do an upload in embrio look in the message log in the bottom right of the screen, it prints the command line used to upload through arduino which includes the location of the compiled code.

The code embrio generates is very different than what you would write yourself, it doesn't even attempt to come close. Embrio works by isolating code into self contained chunks and executing those chunks at regular intervals. Just in philosophy that is very different than how a person would write code, so things will be diffefent at the code level.

One good thing about embrios architecture is that at design time you can have a very good idea of how long your program takes to run. Look into the profiler tab. This shows you how many microseconds each node takes to execute on your hardware, and because we know how often each node executes there is a pretty accurate estimate of how many milliseconds per second the processor will be used. I'm on a phone so can't link it but there's a video about this is the tutorials section.

I'm not sure how an embrio program compares to a hand coded program that accomplishes the same task. That would be interesting to know but it's not an experiment I've done yet.