The Congressional-district method of choosing Presidential electors is the easiest way of getting away from the unit rule (the rule that all electoral votes in the state go to one candidate), because so far as I know (based on my experience in Colorado), that's already how the individual electors are chosen.

Each party has meetings for each district of each type in every election year for the respective office. At the Congressional district meetings, the delegates choose the party's candidate for that district, and in years divisible by 4, they also choose a Presidential elector. Then at the state party convention, the delegates choose the other two electors that the state is entitled to by Article I, Section 2 of the U.S. Constitution. It doesn't have to be done this way, but it's the easiest way to let people in every part of the state have equal input.

Of course, if each state simply chose how to allocate the votes directly, rather than choosing people who then cast the votes, the unit rule could be avoided in any number of ways, one of which was put on the ballot in Colorado in 2000 or 2004 (I forget which).

Until this business with Virginia and Pennsylvania, I would have welcomed a change from the unit rule to the Congressional-district method (don't call it proportional) of choosing Presidential electors. I hadn't realized just how skewed the outcome could be.

I really need to put up the exact wording for my system of objective districting. If someone (nudge, nudge) then publicizes it, Democratic legislators in every state will have something to work from. What I have is specifically written as an amendment to the Colorado constitution, so citations of articles and sections would need to be changed. Moreover, coastal states would have to insert some provision for determining whether a waterway makes two counties non-adjacent.