Neighbours don't really have to agree all that much, especially if we use the variable plots I'm advocating. For one thing one does not expect medieval maps to even be semi accurate. And with the variable plot method the intervening areas between maps will be filled in after having established earlier maps with which they intersect. Here is an example of what I mean.
We are not doing an atlas; we just need the map to be easy to understand and to look good.


Ok I see the point. So the space between plots will be mapped after the surrounding maps are done but who will do it? Is it 50/50 between the two mappers or is it collaboration between them in order to blend both styles together? I think the latter would look better but it is also more complicated.

About mapping a city; I don't think we should lock the plot since the city is probably just a dot on the regional map. But we should lock the city. Also, since city shape depend on the landscape, it might usually be better to map it after the regional map is done but not mandatory.

The plot arrangement does not have to be a grid, you are right but it just need some order. As long as we define specific area (just like you did) it should work fine. Also, I changed my mind about political boundaries; they are not that good for plot placement.