As I’ve suggested elsewhere (in Tutorials), “OpenGL Renders” are your BFF = Best Friend Forever™ when it comes to animation projects. These will produce geometrically accurate outputs, in the same file-formats that are used for “real” renders, and they will do it fast.
(And, with a little fiddling and creativity, they can be the foundation layer of ‘finished’ animations, upon which you add “spice” that is conventionally rendered.)
“Geometrically accurate” means that you can drop-in replace that OpenGL render with a conventionally rendered “final” and it will match, exactly. So, if you plan-ahead about the dimensions and scale of your sets and objects and in the placement of your cameras, you can actually do final-cut editing of the entire show, before final-rendering any(!) part of it. (You even do the first pass of this process before you build final versions of characters, props, and sets! Just keep it all “in scale.”)
You can make a director’s and a cinematographer’s final decisions, so that the only task which remains is to render the specific frames that you (now) know are required, doing each in the most-efficient way possible (which might well be: “OpenGL,” or a hybrid approach that uses OpenGL to do 90% of the work).
Given that an OpenGL frame can be produced in half-a-second whereas a conventional render might take minutes or hours, that is huge. That’s a “game-changing” difference in efficiency of work-flow.
You can make those decisions freely … experimentally … playfully … because they are “free.” That’s important in coming up with a really good show that resembles “real” shows, which are done that way. Getting an image that you can look-at and consider, and decide with, is no longer painful.