Planning for additional capacity in a 3 node cluster to handle a maintenance mode scenario doesn't really make sense.
You would need at least 4 nodes if you wanted to evacuate the data on one host to the remaining hosts in the cluster, and maintain a failures-to-tolerate = 1.
The reason for this is that you need at least 3 nodes for FTT=1.
I would let your team provision across all 3 nodes in this scenario - if you plan to introduce a fourth node at some stage, then it is worth setting that aside for this exact reason.
In fact if you plan to go to 8, 16 or even 32 nodes, and still use FTT=1, having one node worth of capacity is still enough to set aside for this.
If you plan to use FTT=2, consider setting aside 2 nodes, and so on.
HTH
Cormac
http://cormachogan.com