"We don't know how to make our orphan-crushing factories not to crush organs, so we oppose legislation that would require us to operate factories without crushing orphans. But hey, going with the times, how about we offer you Noble-Orphan-Crushing-Peace-Prize?"
I'd say: "If we can't figure it out, then maybe the orphan-crushing factories must close and we need to revert to the times we had before we built orphan-crushing factories."
However, a simple compromise could also be. If you don't have identified yourself, you get limited access - you only get moderated content, you don't get infinite scrolling (but mabe no one should?)- we simply assume you're a minor.
However, it might also be that we don't really need an ID or age verification system if we change regulations on how content needs to be moderated, how it's chosen, how much you are actually in control, and how is it presented (infinite scrolling, threads, pages).
There could also easily be a multi-tiered system - small enough, you don't get that many regulations, because you aren't affecting a large population, large enough, you either implement some age verification and/or ID system (no matter how hard that might be), or you spend more effort on content moderation, use a different content algorithm and/or provide stronger measures against manipulation by bots.
I am not sure that users should be required to provide their name to the public. I think that can have a chilling effect on online discourse and can easily lead to people being intimidated by mobs. (There are cases where people who used their real name on twitter where harassed not just online but also in the real world, I know at least one famous case where someone committed suicide following such harassment).
Anonymous identification is possible - e.g. you can use for example a German ID to anonymously confirm yourself, AFAIK, it's not a commonly used or known feature yet, though (and I think German's ID isn't the only one). And while fake passports, driver's licenses and IDs can be made, I suspect the digital fakes are a bit harder to pull off...
It ensures that the person actually has a proper ID and so creating fake accounts is harder, not saying it's impossible, especially if a government agency were complicit. But for that, maybe the minimum information that needs to be visible would be something like the Id provider (e.g. country, but maybe the regulation doesn't need to limit it to only that. We also have SSL certificate providers whose responsibility is to identify businesses so people can't just open amazon.con and pretend to be the Amazon company, maybe they could also implement an ID system, depending on how the regulation is written).
Companies and organizations already often need to find ways to identify themselves to services like Google or Facebook, to do their advertisements, use APIs or provide apps on their platform. Of course that system isn't quite as convenient as using your personal ID, but that's okay. Companies and organizations don't require anonimity.
