A malware might create a service which opens a previously closed port on your system. An independently configured firewall would keep the port closed, even if the service was running without your knowledge, hopefully blocking whatever activity the malware was trying to do.

Also, you can configure the firewall to drop packets coming in to closed ports, rather than responding to the sending device that the port is closed. This effectively black-holes the incoming traffic, so it looks like there’s just nothing there.

If an attacker already has access to a system, they can use hitherto closed ports to communicate with C2 servers or attack other devices. In that case, a firewall that only allows known-good traffic will prevent further damage.

You can set up an intrusion detection/prevention system, that logs/blocks certain traffic. If you do have public services running, you could block access based on location, lists of known bad actors etc. I guess you could argue that this is beyond the scope of a traditional firewall.

Firewalls set and enforce policy. Closed ports are only incidentally secure. Also, they can do a lot more than answer “nothing is listening here”.

I don’t use a firewall apart from router, but if you set a firewall in all of your devices, the chances of one of them getting infected and spreading it to the others via LAN would be low theoretically.

A modern firewall might also block connections to known bad sites, in case you do somehow get malware reaching out to a command & control server. Or it might identify malicious application traffic over a port that should be for a more trustworthy service.

But these are usually only a concern in places like businesses or schools where there are a lot more people, devices, etc. on the network, especially if there’s a guest network.