Except they're testing to see if I know the specific process they want to teach me.  If' I've got another process where I just think it all through in my head rather than writing it down (I can think far faster than I can write) and still get the right answer, I shouldn't be docked marks for using a different - but still just as accurate - process.
		
		
	 
It doesn't have to be the single specific process they want. It just needs to be a process 
that you have shown to them.
Believe me, I hear you. I do almost all of my math in my head. Even calculus stuff, I usually do in my head. But showing your reasoning--whatever that reasoning might be--is very important.
	
		
	
	
		
		
			And sure, lucky guess will occasionally get me the right answer.  When I get 42 or so right answers out of 50, though, that's way beyond what guessing can get me; I've clearly got a process for getting to the answer, it's just internal and has little if anything to do with what they're trying to teach me.
		
		
	 
Then you should show them that process. Spell it out, tease it out, 
display it. That really, truly is the point. It's sort of like asking high school students to write analytical essays of extremely well-known, well-studied works like Shakespeare. We don't care about whether students know how to structure a five-paragraph essay, that's pointless. Practicing the skills of developing a meaningful claim, finding the evidence to support that claim, and (most importantly of all) 
communicating that evidence to someone else so they can understand what you're saying, are incredibly important skills for anyone.
No one cares if you can prove the Riemann Hypothesis 
only in your head. No one cares if you have an unspoken mental model of quantum gravity. They care about you being able to 
communicate that to other people so 
they can also understand it.
That's why we learn mathematical skills beyond basic arithmetic. Not to have the ability to solve a quadratic equation mentally. It's to have the ability to explain to other people why that, and only that, could be the solution to some problem.
Being factually correct isn't enough. You have to explain why you're factually correct. That's literally what human knowledge is, the ability to show 
why something is factually correct.