Maybe that doesn't matter because posting history is public. One could easily envision another site that had some non-public data associated with email address, pseudonym, and no other PII. That site definitely violates the requirements you describe.
That might be as simple as requiring someone to log into their account (as you can already do on HN) to sending them a confirmation email, or other.
> What if they got a request like this from someone who claimed not to have access to the email they used to sign up? HN couldn't possibly comply.
No, and nor should they.
> One could easily envision another site that had some non-public data associated with email address, pseudonym, and no other PII.
"Dear requester, please log into you account and use the "request my data" link on your account page"
There's a few things written about this around the web:
https://law.stackexchange.com/questions/28998/right-of-acces...
> ISTM the system you describe could lead to more data collection.
It really doesn't. What it does is attempt to gatekeep data (which these days is a huge risk) collection to organisations that are competent. You always have the option of not collecting data if you don't want to take that responsibility.
They wouldn't have to. If you're requesting your data from a company, you have to burden of proof that you are in fact the person about whom you're requesting the data. If you cannot prove it, you cannot get the data.
This is an insanely hard problem to do well. The recognition of customers is literally through which phishing attacks work. This is one of the most common avenues that people get access to systems and data they shouldn't. GDPR mandates that virtually every site needs to deal with this problem now.
Also, I'd say that reprieving all data on someone is not a trivial problem. Imagine you had a forum and somebody requests all of the data on them. However, another user has made a copy elsewhere on the forum with some notes about that user's personal data. Does the requester get access to that copy?
And now keep in mind that you complying is based on your understanding of a regulation that companies have spent millions on lawyers to figure out and still aren't entirely sure. And you have to follow this regardless how big or small your website or business is.
I do see your point though, the intersection with phishing attacks is not something I'd considered in the context of GDPR.
However, as long as you provide a way for users with an account to download their data, you're 90% of the way towards compliance.
Again, unless you are FB or Google, the regulators will cut you some slack (just look at the lack of bankruptcies caused by GDPR).
Really if you can't established a way to recognise customers how can we (as customers) expect you to competently do 2FA and competently store data without losing it.
It's like complaining that your kitchen shouldn't have to keep track of expiry dates of meat; if that's too onerous for you then you're showing you're not competent to run a kitchen.