The activity is the point, not the object it leaves behind.

Everything valuable happens while the hand is moving, and almost none of it survives on the paper.

Watch a young child work and try to describe what is occupying them. Almost all of it is physical. They are learning what a hand loaded with wet colour does compared with a dry one. They are finding the point at which a crayon skids rather than marks. They are discovering that a shape can be filled in and that filling in is deeply satisfying and takes a long time. They are working out that their arm can swing from the shoulder and make a long arc, or pivot at the wrist and make a small one. This is genuine research, conducted seriously, and the child is entirely absorbed in it because the answers are new.

None of that research is legible afterwards. The sheet shows a smear and a tangle. An adult who only ever sees the sheet concludes that not much happened, and starts looking for the day when something recognisable appears so that the activity can be judged a success. That is the wrong measure applied to the wrong object. If you want to know whether the session went well, the evidence is in the twenty minutes, not the artefact: how long they stayed with it, how deep the absorption went, whether they tried something they had not tried before, whether they went back and did it again.

Holding this in mind changes practical decisions. You stop rationing paper on the grounds that the last four sheets did not produce anything worth keeping, because the sheets were never the point. You stop steering towards subjects that will yield a nice result. You stop timing sessions to fit an outcome. You put out the materials, you make the time genuinely available, and you accept that some days the entire session consists of mixing every colour together until it goes grey, which is a legitimate and thorough experiment with a reliable result that the child needed to obtain personally.