We can, for instance, add genes to plants to produce substances that bugs don't like and we can eliminate genes from animals that inhibit growth. It's not insurmountable to edit genes to "do more" or "do less"—if the answer is making more or making less of a protein we already know about. But changing the design of something is a hard challenge that we aren't good at. If increasing VO2 capacity was something that could have evolved as a beneficial mutation by way of increasing or decreasing the presence of a protein, it probably would have happened. But the fact that it hasn't either means increased VO2 capacity isn't great (eliminating muscle growth inhibition, as it turns out, isn't great) or requires more than just tweaking a protein amount.
Which is to say, gene editing is the ability to change the code. It requires separate capabilities to know what to change it to. This news is great for areas of research like cancer treatment (imagine taking a puff from an inhaler to cure your cancer, which is just ~2 gene mutations!) but much less exciting for "enhancements".