Two ways of doing this. Easy way (and the way that NI tells you to do it): create a control template for the cluster. When the template changes, everything that subscribes to the template changes accordingly. The limitation is that it is still a cluster and those aren't the easiest things to work with. Your circuit analogy is a good one in this case.
Harder way (but worth it and something that NI doesn't tell you): Store all variables as a 2D array of either strings or variants. For each row in the table, the first element is the variable name, the second is the value, the third is the data type and the fourth is the units.
This is a royal pain in the *** to code initially, but will save you limitless time later on as you end up getting things like state persistence, search-sort, type conversion, and so forth.
Over the decades (yes, I've been coding LabVIEW for a couple of 'em), I've come up with an entire suite of VIs to work with these variables in 2D arrays just as though they were named members of clusters.
Don't know if the paper and sample code is still up on ni.com, but I did a presentation on this at NIWeek 2006. Hope this makes your LabVIEWing much more pleasant...