What do you have in mind?
Does avoiding the ACA subsidy tax cliff really save much? That's the question that I think is important to have answered. Currently, a cliff-on vs. cliff-off i-orp run isn't quite apples to apples because a cliff-on run has a significant reduction in taxes for model years below age 65 that is not credited in the model. For me, $11K for 9 years. For people who will work to age 65 or those who can't qualify for the ACA, this feature would be of no value. But for those of us with years of ACA in our future, it would allow us to see if we really should bother with cliff avoidance.
From an implementation standpoint, the idea would be to have another input: "Annual ACA Premium Tax Credit" (PTC). This would be used if the cliff avoidance option was selected. This field would be defined as the number of thousands of dollars annually that the family would get in tax credits associated with having purchased health insurance from one of the exchanges. Or more succinctly, the number of thousands expected on line 24 of IRS form 8962. That value, indexed by inflation, would reduce the federal income tax for model row ages less than 65 (the ACA doesn't apply to Medicare users of 65 and older). It's a straight reduction in taxes, even allowing taxes to "go negative", meaning the government pays the taxpayer.
Now, the edge cases. These, I think, can be managed or ignored.
What happens when one member of the couple turns 65, goes on Medicare and the other is still using ACA? I suspect that the PTC could be halved and things would work out about right. An alternative would be to have a PTC input value for each spouse. That would be more complicated for the end-user to complete the two fields for most couples, since with the typical a joint policy, the number is not broken out by spouse. The only time it would make a big difference, I think, is if there was one person who was much more expensive to insure than the other (a lot older and/or a smoker). But I'd say halving it would be 95% of the way there, and not to bother with and input for each spouse.
Another edge case would be hot to handle a family size change. The PTC goes down as the kids leave the nest. I'd say this case can be ignored without compromising much; the expected PTC can be set to the forecast for the PTC value of whatever the majority of years will be, or some kind of average determined by the user.
I'm not sure how other i-orp users feel about it, but I think it would be cool to put a tidy dollar figure on what the PTC is worth by comparing one run with, and one run without doing the ACA dance.