Because cancerous cells prioritize their own growth over the health of the wider system in which they exist.
Companies with this mindset do not reach a point at which they say "ok, we're big enough now". There is no target size they're trying to reach, there is only a target year-over-year multiplicative growth rate. They prioritize growth over providing value to the users they have. They grow without regard for the effect it has on social dynamics within the system (changing what the product is good for and what it's bad for), or the effect it has on social dynamics beyond the system.
It's the Paperclip Maximizer approach to business. And, of course, it's the stock market approach to business too.
They want exponential growth purely because it is exponential. They aren't chasing anything other than that metric, they didn't ask how they could make it better, they didn't ask their current users what could make it better, they aren't asking if they are creating value, they simply said "huh, 400mil isn't enough, give me more". It's the obsession that makes it pathological.
> they didn't ask their current users what could make it better
They often do, this is part of the innovators dilemma in that early adopters aren't often mainstream customers, so the feel unheard, when in fact the company is listening to its more mainstream users (once they arrive) - which often causes the early adopters to lament and chase the next thing.
> They aren't chasing anything other than that metric
It isn't that these companies aren't listening users, or aren't creating value. They would fail if that was the case. The metric is profit and a big presumption is that with free-trade and free-markets any transaction is mutually beneficial. If they don't provide value to the customers, the customers don't pay. If they keep providing value for more people, enough value that people will pay, then value really is being created.
I think the concerns more have to do with with:
* Hidden costs - such as giving up your data or privacy aren't usually clear (and if they are, in the sense you are giving your data, the implications and risks aren't clear)
* The value might only be "on-net" better. For example, network effects have "value" in the context of the data, even if better platform might exists that don't the enough critical mass participation.
* A related problem to the above, the product might be on-net better but still not optimal for your country, demographic, social group etc ... (non-bay-area-based-white-affulate-males)
* Markets aren't always free (enough) so disparities and distortions cause real issues. All sorts of externalities abound.
* Indirect user/purchaser models like advertising will optimize for things that while they may maximize overall profit (if done well) are NOT optimal from a purely user-centric view.
* Change, of any kind for any reason, good or bad, is often lamented by users
