So is it not possible that holding tools - or rocks to throw at adversaries, etc - would have been immediately something that an ape down on the ground would frequently do, with no further evolutionary step required to enable it?
The ape carrying tools around is most often - by far - using one hand to carry them. That same ape on the ground is unlikely to handicap itself by walking on its hind legs for any distance, with or without tools; when traveling it will carry the tools, rocks, etc, with one hand or in its mouth, because three legged locomotion is much faster and much more efficient than two legged for quadrupeds. It will stumble around on its hind legs in specialized and temporary and occasional circumstances, but not as its primary mode of locomotion or transport - the energy cost alone would be prohibitive, given the absence of any immediate payoff, and when the extra risk of predation and injury and hampered food acquisition and so forth is figured in its hard to imagine why any animal coming down from the trees would do that (and easy to explain why none have, outside of the presumed hominid outlier. Bipedalism has emerged only from the ground up, only among small animals, and other than the slow and awkward but heavily defended scaled pangolins only in jumping animals. There are no ground-immigrant bipedal squirrels, raccoons, weasels, monkeys, lemurs, baboons, etc. The apes that we know came down from the trees unto the savannah became better land quadrupeds, not bipeds, of course: that's the obvious way to go. Climbing animals faced with an incrementally more significant grounded life are not going to begin adjusting by crippling their ability to run and their ability to climb and their ability to hide in one move.
Remember that the bipedal transition happened in a quadrupedal ape. Sophisticated employments of necessarily free hands necessarily came after, not before, those free hands were available. The earliest, chimp-brained, crude knapping, toolmaking forerunners we know of (including the carrier bag employers, the probable first hominid tool) were already obligate bipedal primates - hundreds of thousands of years of major skeletal alteration via selection pressure down the evolutionary path to us. So what in hell could that selection pressure have been? Seeing over tall grass? C'mon - that's not even a serious Just So Story. (And it doesn't match the timeline).
To abet the evolution of bipedalism one needs some kind of selection pressure strong enough to overcome the obvious and very large penalties of extended bipedal behavior in a fairly small quadrupedal monkey-brained animal well-adapted to climbing trees. These include a reduced, not enhanced, ability to carry objects long distances overland . They include a reduction in flight speed, evasion agility, and concealment, from predators. They include greater exposure to injury, and risky weaknesses in other biological systems of the organism (circulatory, digestive, motor control, etc). The grounded quadrupedal tree-adapted primate on its hind legs can't run, hide, or fight - and it frequently injures itself just walking around. Even today, after a million years of adjustment, healthy and athletic humans sometimes get dizzy
from standing up. A thirty pound weaponless version of that in the crosshairs of a stalking leopard is lunch.
The origins of logic begin by presuming logic has originated, after all. That presumes reasoning - the larger and inclusive context in which logic is embedded - is well established. One hates to be too cynical, but "seeing over tall grass" came from the topflight pros and experts in the field.