Haven't noticed that before, must be relatively new. Remote ID is what the remote system uses to identify itself. This varies between IKEv1 and IKEv2 - but lets stick with IKEv1.
The most commonly used option is "IP Address" - being the IP address configured on the remote devices WAN interface. If the remote device is behind another device doing NAT then the peer IP address is usually the public IP on the outside of the NAT device and the remote ID is the private IP on the outside of the VPN device.
Another option is "dn" for distinguished name. It is used during certificate authentication. So not applicable in Meraki world.
"hostname" is another option. This can be popular when the remote site is using a dynamic IP addresses.
Some systems support a "key id" for selecting a key.
Some vendors probably have some other options.
Although I said lets ignore IKEv2, IKEv2 also added fqdn and email address as ID options. I find these much easier to work with. Every company has an FQDN available to them, and you can easily create dummy email addresses (they don't have to be actual real email addresses) to identify sites, suchs site1@company.com, site2@company.com, etc.
Now you can create much more scalable VPN solutions without having to configure the IP address for each remote branch on the head end.
But if you have AutoVPN - you don't have that issue anyway.