WireGuard doesn't do any forwarding out of the box, you need to set up your iptables/nftables to get all of that working. If you follow the WG quick-start guides, they often work by masquerading traffic, making VPN clients act the same way a bunch of computers behind a NAT router would.
You'll need to set up all other kinds of routing as well for cross-network discovery to work. WireGuard doesn't do broadcasting in general (it's a point-to-point protocol after all) so you'll need to wrap broadcasting protocols manually.
Other VPNs go more low-level (at least in TAP mode), mirroring an ethernet network with all the broadcasting and low-level protocols you can think of. In theory you could do that in WireGuard (running L2TP over a WireGuard link) but many phones won't support that, and it'd probably be just as easy to set up an OpenVPN/IPSec+L2TP VPN in that case.
I'm not sure if it's a good idea, though. I imagine most people wouldn't want a printer publishing its mDNS hostname to wake the 5G radio on their phone, or the battery level of their laptop in the case of KDE connect.