There's different ideas, roughly distinguished between market and council-based, with less and more central planning. The CNT in Spain had a quite market-based approach, for example, but OTOH you see council-type structures even in today's capitalism: The development of lithography technology and machines for that, for example, quickly hit a brick wall, none of the (gigantic) companies working on it were actually large enough to do it so they started cooperating. The label on the machine might say "ASML" but really they're only a systems integrator: They've worked with a multitude of other companies to develop and build exactly the stuff that will be necessary, there's no competition between say Corning and Zeiss who's going to make a particular lens or such: They've agreed, together, to build a certain technology, divided up the work according to their specialities -- including "make money with the machine", that's TSMC's area of expertise. Roughly speaking: The less commodified a particular erm commodity is the more likely it's not actually directly bound by market forces, even in our current economy you get these islands of horizontal cooperation within the larger shark tank. You'll pay money for those machines but money alone might not buy you one, you might need to be part of a syndicate.

But I agree with you (or I think that's your implication) that pure mutualism will not work for these kinds of "put a man on the moon" projects, it's not structured enough and without structure no planning (centralised or decentralised). And frankly speaking the theory around this topic is kinda lacking, first off because much of the theory about it is old, where "big industry" meant "a steel mill", secondly because Anarchism has ceased to plan ahead details: We don't have the necessary knowledge and information to pre-empt the decisions of people down the line, and we shouldn't attempt to, either. They will organise those projects as they see fit in some democratic manner, what's up to us is to grow democracy within the economy to a degree where more and more economic actors jump on the ship, as well as develop abstract frameworks, a body of ideas and approaches in line with Anarchist principles, that they can pick and choose from as they like and circumstances dictate, and develop further. And most of all we need to kill off hierarchical realism, that is, the idea that nothing ever works without an imposed hierarchy even though everyone sees it working all the time when friends get together to have a grill party. Are there scaling issues, sure... but hierarchies have scaling issues, too, even insurmountable ones (mostly around information processing complexity and perverse incentives) and we don't discount them on that basis. There's a strong cultural bias and blind-spot, there.

In a nutshell, it's the old leftist problem: We know exactly what's wrong and also know how things ought to look like to be better, but details, man, details. In the end, in practice, there's no perfect, there's only less bad.