Rather than Macallan 12 which is sherry cask matured, as the sherry has a dominating effect, I'd assume a longer finish would be necessary to see much of an effect - depending on the finishing barrel and that your barrel will be finished with rum (as opposed to tequila), I'd expect more synergy thus a longer finish time required. Thus, to see more of a rum impact earlier, I'd start with a base whisky that is lighter and having less barrel impact than something like sherry. Start with a first fill (ideally) bourbon cask matured whisky as the rum impact on the whisky should be more obvious than on a Macallan 12. If you're set on Macallan, I'd do the Macallan 10 fine oak. As this is somewhat of a heuristic experiment, you might save a little money and try this on something like Speyburn 10.
Since your cask size is going to require more than one bottle to fill, the great thing is you can draw off samples at different ages and keep them for comparing later. If you tasted and drew samples off at 4, 8, 12, 16, 20 and 24 months, or something, you can see what impact the rum is having overtime and then determine where you like the impact best. Of course, if you love it a 12 months and hate it at 16, you may have the remainder of the cask filled with something you despise. That's the risk. But assuming you get to a point of having 3-4 different age-lengths of finish, when comparing all of them at once you have a nice start to an evening!
Interesting comment blenderm makes about no law requiring statement for how long the "finish" is. Essentially, any whisky that is aged in more than one cask is "finished." Balvenie double wood is an example most people have no problem with. But I see his point, and it's a good one.