Do you have a source for that claim? How does sharing IP's with 'like factories' and thousands of other legitimate users justifies getting captcha-blocked by Cloudflare without explicit instructions from the website owners?
If the like factories are using 3rd world IP's they are probably real human ( https://www.rt.com/viral/388169-smartphones-factory-generate... ) because bots can be run cheaper on a 1st world server because bandwidth, IP, and power usually cost less there. Captcha is an anti-bot measure and is not (or at least shouldn't be but I'm finding it hard to pass CF's captcha as human) very effective against actual humans.
I don't even think CF considers(or even aware of) the existence of a like factory on the same network for displaying the captcha and I am pretty sure they don't justify their captcha blocking with it. It's more likely that they just see a bunch of connections coming from the same IP and naively concludes that it must be a bot.
I'm sure it's easy to make "excuses"(vs justification) but given the real harm it does to actual human users and questionable effectiveness against the doubtful ill-effects of the existence of like factories against CF-hosted websites; I'd like to hear that "justification."
First world server IP ranges are treated just as poorly for this exact reason. Try browsing the web through a VPN/proxy on an OVH (French) server. I get captcha requests on Youtube videos and Google searches.