A standard example of such visualisation is 'in bluish-white light on a dark background, imagine a line. from the same light, imagine drawing a circle using the line as a diameter. Now draw an equilateral triangle touching the boundary of the circle. Then a square around the triangle. Finally, draw another circle around the square, and hold the resulting image stable in your mind' (yes, this shape probably inspired Rowling's description of the deathly Hallows. It's a very old exercise).
I can visualise the line and the first circle easily. I can add the triangle, but at this point it already takes some effort, I can't do it if I'm even a little distracted. Typically, adding the square makes me 'lose' an earlier detail, usually the line, from my visualization.
That's probably not very informative to your question though, because visualization doesn't have the same sorts of rules around complexity a graphics screen does, the brain definitely uses some form of internal shorthand. I can picture a particular celebrity as easily as I can picture a circle, and picturing her holding an item from my grocery list in each hand is if anything somewhat easier than picturing two other elementary shapes connected to the circle. Maybe because humans are used to imagining other humans holding things, I'm not sure.
In any case, there are definitely limits to what a person can visualize, and they're pretty modest for the untrained mind, but it's quite unlike a graphics program where one could simply count the number of pixels required. Not sure if that helps or not...