And in the good old days almost nobody was trolled because the energy required to send a message via snail mail (starting with "do I even own stamps any longer"?) is several orders of magnitude higher than composing a one sentence death-threat on twitter.

It's probably a crime, yes. Unfortunately, you'll be prioritized accordingly. When prioritizing a crime that involves some anonymous troll responding with a death-threat or threat of harm that was most likely done in a moment of stupidity/weakness rather than with an actual intent to commit a crime, you're going to show up pretty low against the list of rapes, murders, assaults and thefts that are currently being handled. Once you've identified that the threat is such that it actually is a crime and can be successfully prosecuted, you have to go through the trouble of finding out who actually left that threat. Should they successfully uncover that information, the likely penalty might be so minimal that it's not even worth pursuing. Because of all of those circumstances, all but the most extreme cases of these get completely ignored (and many of the extreme ones).

And we can fix that problem by using wide-reaching and invasive surveillance that allows the police to de-anonymize anyone once they get the go-ahead.

But that's probably (definitely) a terrible thing to do, so maybe it's better to find other ways to solve these problems.

Parent and me were talking about hardcore cases. There is no technological solution for people stubbornly following you everywhere and sending death threats.

Furthermore, from some anecdotal evidence, I think most victims in such cases already know who is harassing them.