You’re not correct. Inspection time for a 3x3 solve allows you to plan the early moves, but no farther. It is possible to solve the entire cube in your head (i.e., blindfold solving), but not in 10 seconds. Progress through a solve goes in several stages, depending on your method. Each stage may require several sub-stages.

E.g., for the CFOP method the second stage is pairing middle edge pieces with bottom corner pieces, and putting them in place. There are 4 of these middle edge pieces, so that stage has 4 steps. Each step may require from zero to ~8 moves.

Speed solvers typically scan the cube for the next step, and then while their fingers execute the moves they scan the cube, looking to setup the next step. Sometimes this allows them to anticipate the next _next_ step.

The fastest blindfold solvers in the world have their memorization under 10s. It encodes all the information they need to solve the cube, so arguably they are able to solve the entire cube in their head in under 10s. Execution takes longer, though, because blindfold methods at that level are still less efficient than traditional methods. However, one of the best BLD solvers, Jack Cai, has used blind solving during a normal 3x3 event, by doing his memorization during inspection (inspection time is included in the total time for the actual blindfold event). Then, with a blindfold on, executes the solve. He had a sub-12s average, IIRC.

>but if I see these insanely fast movements in the solving phase

As an old fart who was a kid when the Rubik's Cube was introduced to America, I'd like these young whippersnappers to solve on the original Cubes, not these Ferrari cubes that never stick or get filled with hand salsa and backpack detritus.

Basically most methods of solving go from more intuitive phases to more algorithmic phases. Cubers plan the intuitive phase earlier on in inspection, but the later phases is mostly pattern recognition. So there is still decision making but less complex ones than are necessary at the start.

There's not really any decision making in solving the cube when you have learned the method, you just see the pattern and execute the algorithm. That doesn't really change at all with the inspection time.

There is still plenty of decision making. For any scramble, there are usually a few dozen ways that make sense to make a cross. A top cuber may choose a slightly slower way to do the cross to make the first f2l pair easier (or even solve the first pair at the same time as the cross). They may choose an unusual way to insert the second f2l pair to make the third pair easier. There are a handful of videos out there of cubers breaking down their solves and explaining their decisions.

The last layer doesn't really have any decision making though.