What's the most efficient way to have the same locomotion state machine logic in a TPS for multiple weapon types (no weapon, rifle, sword, etc)?

One way is to have multiple state machines and switch between them as needed. For example, one state machine for no weapon, one for rifle, one for sword, etc. Although this is the method I see most commonly on youtube, it seems inefficient as all state machines would be nearly identical except for the animations specified within the states. If I find a problem in one state machine, I have to remember to go fix it in all of them.
Another way is to have one state machine and where you specify the animation, have a “blend poses by bool” to pick the right one. One downside is that if you have several weapon types, it’s probably going to get messy in those states; Especially since I don’t see a switch or select node I could use, so I would have to daisy chain some blend poses by bool nodes.