I think the schedules laid out in Nicolaides' Natural Way to Draw illustrate what you need to break through a plateau: varied struggles. Even if you think you're terrible at faces, don't work just on faces; Even if you think you are terrible at pixel art, don't work only on pixel art. Cycle through both "types of study" and "mediums", giving all of them equally serious attention; This breaks you out of the dogmas you develop when working mainly in one particular medium, on one particular kind of subject.
1a. Tangents. In pixel art it's easy to pile these up, because you really only can do 0°, 45°, and 90° intersections. Tangents encompasses a lot (lines, intersections between clusters, spaced relations between clusters as in dithering or banding). In art in general, I simply was obsessed with going fast, without proper preparation.
1b. I have to say I just pixel whenever, not regularly (as opposed to drawing, which I do every day). To me, the pixel aspect is minor - pixel art is functionally just another (painting) medium which happens to place an extreme emphasis on design. As I understand it, that attitude is partially a function of pure volume of experience (and experience in other mediums was critical here, IMO), and partially spending time analyzing it in a formal way (eg. mathematical basis of AA, formal encoding of line segment patterns).
Maybe it'll help to say that all the optimizations of the picture I do in pixel art are the same ones I do when drawing or painting on paper -- they're the same principles, stretched to the extremes. Pixel art was well suited to teaching me certain of these artistic principles (eg. conservation of detail, strategic exaggeration, painting order), and unsuited to teaching me others (eg. perspective, tangency, flow)
2. The illusion of color. I say illusion, because when I was starting -- with EGA color -- you certainly couldn't just pick the color you wanted and lay it down. The illusion of color isn't expendable today, but it's more marginal (you still need to think of your picture's colors as one whole thing containing relationships, but you can select the main colors reasonably precisely even if stuck with a fixed palette).
I don't think there is anything expendable today, actually -- art is like this all-devouring blob, if you learn something there is probably a way to apply it to art and art to it.
3. Again, tangents -- there are many levels of tangency you can get tangled in, avoiding most of them is a rare sign of paying proper attention and having a workflow that enables you to be thorough. Also framing/composition -- it's easy to tell the difference between a framing that knows it's telling a story, and a framing that's just like "I needed a canvas to put this thing in".
These both fall in the general category of 'did your homework, acted with clear intent, held firmly onto that intent until the end', which is how I identify .. people I consider genuinely competent.
(I avoid the terms 'experts' or 'professionals' because I think the meaning is usually unclear. I think that you were aiming for 'people of notable competence' here.)
There is another particular concrete skill that is really important IMO, but it's hard to tell if anyone has it unless you are looking over their shoulder when they are drawing, and it's also pretty difficult to explain.