To my knowledge, you can install a one-node cluster within Windows, and then add servers to the cluster as they come in.
However, I believe you will need to re-run the SQL Server installation in order to make it a failover cluster. The reasoning for this is that the SQL binaries will need to be installed on that secondary node, and the services will need to be configured properly.
I have not personally come across this, but I can propose how I would approach it.
One way to handle this would be to install one instance at the initial install (make this a named instance, SERVERNAME\INST2), and then install the default instance during the second install.
The logic there would be to install 1 instance for use now, then migrate into the 2nd instance (which is installed as a failover cluster) after the new server is added to the cluster.
In order to install the 2nd instance on the cluster, additional LUN's will be needed.
The final SAN config that I would recommend is this:
1 LUN for each of the following:
1. system DB data and logs (master, model, msdb, SQL error log) (25GB, my SAN team only presents LUN's in 25GB)
2. temp DB data (50GB)
3. temp DB log (25GB)
4. user DB data (100GB, varies)
5. user DB logs (50GB, varies)
6. backups (sometimes use local disk for backups)
7. MS DTC (25GB)
8. Quorum (25GB)
You'll need to divide up the disk so much as to leave enough to install the second instance and move everything over, or have new LUN's presented by your SAN admin to facilitate the migration to the second instance.
In my example, you'll need a total of 300GB for the initial instance, then another 250GB to make a parallel installation (the MS DTC and Quorum would remain).
Then, you migrate all DB's and logins from Inst2 to the default instance. After that is complete, you can uninstal Inst2.
I would be intrigued if someone posted who has added a node to a cluster successfully before to know how to get around my logic and save time.
In any event, I hope that makes sense and gives a little guidance.
Thanks,
Steve