If I knew why, I wouldn't believe as I do. I wouldn't spend all my free time on forums like these, or doing interviews, or fighting the good fight. I'd be content to be right with the popular opinion.
I choose the non-euclidean map then, if it doesn't matter which flat earth map I use. Wilmore's map isn't a bad second choice. Other choices include the south centric flat earth map, the collapsing flat earth map, and the Greenville map.
I would put them with the receivers and transmitters. If I had an unlimited budget, I'd likely use pseudolites (you can find detailed engineering descriptions to these in the IEEE XPLOR database) to increase coverage. I'd probably use these in a caching sort of way, and be certain of the time it takes for them to talk to the system of record.