this is a very nice and delicate point to touch.
Now as it is, the system is somewhat stupid. I see the world as if there is an overmind in Wien and all the governors brains are somehow astrally tied to it for me to see. For example I am playing as Austria, France DoWs me, I pause the game and immediately send my troops from 20 different provinces in alaska, in tierra del fuego, in indonesia, in siberia... to 20 different provinces of France (or her allies') all around the world. In the end all the hopeless TP's of the AI alliance is burnt down to ground and a lot of colonies are lost. This way I cannot feel satisfaction because I feel like using satellite communications in 1600s.
Now this is nonsense.
On the other hand, giving away the control of the armies to the appointed generals (ie AI), or acting like the governors-generals of the era has no self-sufficiency hence waiting for a little messenger going from London to sacremento is as nonsense as above.
My opinion about this subject is as follows (a little complicated first but very simple indeed):
All provinces have a "view" of the world with a lag. For example I am portugal in 1 jan 1650 and I took Toledo from Spain 4 months ago. my colony in Goa in January 1 1650 will see my capital Tago as of January 1 1649 and will not recognize Toledo as a province of me because it will be taken in september 1649 and that time frame has not "arrived" yet to india. When I click on Goa or a unit on Goa the world map "as we know it" will change and Toledo will become yellow instead of green. The "default view" which is how we see the map on EUII would be accessed by one key (like "home") thus enabling us to see our empire alltogether.
A province would have 2 ways of updating its view:
1- news coming from the capital with some delay. For example in jan 1650 Goa "knows" what the capitel knew in jan 1649. And in jan 1649 the capital knew the province Guatemala of June 1648. Thus, in jan 1649, Goa knows the province of Guatemala as how it was in June 1648. This speed should be close to a unit moving from the source province to the destination province.
This delay is a little bit longer if coming from a province of a country which is an ally/vassal/RM. No information comes from other countries provinces.
2- news coming from surrounding environment. This means, Goa will not wait for the information go to first Tago and come back again to understand Vijayanagar lost a province to the Mughal empire. This spreading of information should take much longer than incoming news from territory.
As a result, every province of your empire will collect information about its surrounding with a slow pace, and send it to the capital with a fast pace. Than the information will be distributed to other provinces again with the fast pace. The default view will be the "overmind view" with all the map in its current position. But to make an action (move an army, attack a province etc) the "view" of the province that the army stands on should be used.
For armies on neutral territory or enemy territory, the fast pace should continue to reach the army via the nearest province.
The information spreading speed should be tied to naval-land slider. Naval spreading speed is faster if you are a naval country and vice versa.
For fleets, the "view" of the port province should be used and information should be spread from the nearest port province. The same way around, when armies or fleets "see" something that information should be spread back to the nearest controlled province and then back to the capital.
This is actually a kind of a "dynamic fog of war" instead of a static one. The point is; When France DoWs me Austria, they can attack to any of my provinces because they were "prepared" for a war and knew when it was to start. But as Austria, my provinces should "know" about the war when they are attacked or if they are not attacked, when they are aware of that information.
An addition to this might be this: According to the Centralization slider, every nation has the ability to put a limited number of "command centers" to any desired province. A fully centralized country can only have the capital as an information flow centre. But a fully decentralized country can have up to 10 centers of information (can be placed once, and can only be changed its place lets say every 20 years) that can be anywhere in the country. This means for example France DoWs on me Austria. I have a colony in India bordering theirs. I also have a colony in Jakarta bordering theirs. I have centralization 5 so I can have as much as 5 command center provinces. Lets say one is in jakarta and another in India. I can attack with my armies in Europe because there already "knows" about the war but I cannot attack with the ones in India or Jakarta. But France can do it because they are the aggressor side and they "planned the war ahead". They have an army in India so they attack to my province in india. Now I can use my army in India too because they are now "aware" of the war. There is no need to wait for the information coming from Wien. Then we can assume the governor-commander there is using its will for reality purposes. My colony in Table still does not know about the war but the colony in jakarta does not have to wait for the information come from all the way from Wien to Jakarta. Because now the information is available in India, it starts to propagate from there too. Information of war from India of course reaches Jakarta sooner than the information of war from Wien. Then I can use my armies there to attack French territory neighboring Jakarta. No attacks are being made to Table so the governor there still doesnt know about the war. My colony Table learns about the war as soon as a messenger arrives there. It may be the one from Wien when started its trip when the war is declared or it may be the one from India (which is closer geographically) when there was attacked weeks after the DoW.
I am waiting for any comments about this system.
PS: I am sorry for any grammatical mistakes above, as English is not my native language.