It's been speculated that ReplyGuy is a dev or other non-malicious actor attempting to force nostr to harden and become stronger.

Whether this hypothesis is right or not, regardless, most of the attempts to deal with the spam are using techniques from fiat/centralized social media.  Filtering, blocking, and so on.

None of this is going to work, at least in the long term.  The incentives need to be correct.  ReplyGuy can essentially create unlimited npubs to send the spam from and it's a huge game of whack-a-mole.  Even if the relays all manage to block the replies, new relays will be made that also have to block in the future.

What if sending a note or reply costs 1 sat?  A user can still filter those replies, but receiving one sat each time the spam is sent imposes a cost.  Let's see how long ReplyGuy can stay liquid when there's a real-world cost to spamming the network.

Or perhaps there's another approach?