I think it's more interesting to have different maps. There is however a problem, which is that for instance the upcoming seasonal finals uses the same map pool as WCS EU (despite being hosted in Korea), therefore possibly giving European players an advantage.
So you might want to have parity in the map pool for the seasonal finals. The definition for this would be that regardless of region of origin there are the same number of familiar maps for the global tournament. This assumes three regions, a global map pool based on taking maps from the different regional tournaments and a resulting four map pools of equal size.
This isn't always possible however. For instance, if three regions had the pools A B, A C, B C, then you can't pick a fair two maps. The only way around this is by introducing new ones, or by expanding the map pool to A B C, but this conflicts with the assumptions and therefore won't be considered. So this shows that you have to be careful in the choice of your regional map pools and that some degree of coordination is necessary.
I found out that you can use the following method to find out if the map pools allow you to have a fair seasonal final. What you basically do is line-up the pools per region. You start out with the maps shared across three regions. Then you take the maps shared across two regions (with a third 'random' map added from the region you left out). And then finally the maps that are unique for all regions. Now you give out scores, three points for the first, two points for the second, one point for the third. If the total score is divisible by three your map pool can have parity for the seasonal finals.
I added an example for the current WCS seasons here:
As you can see it's divisible by three so therefore it should be possible to have a fair map pool unlike the one for the seasonal finals. In this example they could replace Newkirk with Atlas in the seasonal finals and achieve this.
I only dabble in math occasionally, so please tell me if I overlooked something.