The true strength is in the open interfaces and common protocols that enable competition and choice, followed by the free-to-use libraries that establish a foundation upon which we can build and iterate. This helps us to stay in control of our hardware, our data, and our destiny.
Practically speaking, there is often more value in releasing something as free software than there is to commercialising it or otherwise tightly controlling the source code... and for these smaller tools and libraries it is especially the case.
Many bigger projects (eg. linux kernel, firefox, kubernetes, apache*) help set the direction of entire industries, building new opportunities as they go, thanks to the standardization that comes from their popularity.
It's also a reason why many companies release software as open source too, especially in the early days, establishing themselves as THE leader...for a while at least (eg. Docker Inc, Hashicorp).
It typically takes a small core team to build the framework/architecture that enables many others to contribute meaningfully.
Most OSS projects get bugger all contributions from outside the initial core team, having limited ability to onboard people. The biggest and most active (out of necessity or by design) have a contribution friendly software architecture and process, and often deliberately organized communities (eg. K8S & CNCF) or major corporate sponsors filling the role.
Free Software and resulting ecosystems seem to have a better chance of contributing to the common good over the long term. This is simply because most companies are beholden to their shareholders, and at some point the urge to squeeze every last cent out of an opportunity comes to the forefront, and many initially well intentioned efforts get poisoned.
Free Software licenses like the GPL help to protect our freedom and to set open standards, and are essential for the core technology stack.
When someone can get annoyed with some shitty software or its license-terms and reimplement the core functionality in a few days/weeks/months ... eventually someone will get annoyed and create some decent free software that will kill off the shitty alternatives, or even just a better commercial alternative. This only works because of the open platforms & protocols.
One of the major challenges for consumers is finding good software today in the grey goo of projects and appstores. This harks back to OP's point about curated collections of software. It's also where the various foundations add value (CNCF, Linux Foundation, Apache) ... along with "awesome X" gitlab repos, which are far better than random youtube videos or ad-riddled blogs or magazine articles.