In general it’s best to focus on trade nodes that have a) high potential value b) you can heavily influence trade there. So, to use my Prussia/ Germany game for example, Lubeck is a trade node with high value, as was Antwerp. Choosing to build in either one was almost always a worthwhile option. I had significant holdings in both, and any increase would represent a decent value. I also had significant holdings in Novgorod, Baltic Sea, White Sea, and Wein (others too, but point is same). Now of those I had a near total domination in the Baltic, but would build to keep Denmark down, since it fed to my Lubeck port. Wein also fed to Lubeck, but was not worth pursuing. It had decent value, but was far too contested and represented a worse investment than elsewhere. Novgorod was 50/50 between me and Russia, but the value wasn’t that high. The White Sea I easily could have dominated, but it fed to the North Sea where Norwegian, English, Burgundian, and Dutch power was so high to make this lucrative node a poor investment.
So from that example Antwerp and Lubeck were top priority investments, the Baltic was an investment that ratcheted down as I destroyed Sweden, but still valuable, and the rest were either too poor or too contested to make the building investment pay off.
That’s not quite how it works. Trade nodes have local production (trade goods produced within that node’s region) and incoming trade. These two add together to create the total trade value in a node. From that you determine the total trade power, and what percentage is collecting and what percentage is forwarding. For ease I’ll use an example Baltic Sea.
Lets say that the Baltic Sea has 30 local production, and has 20 incoming from Novgorod and Krakow. This means the total trade value is 50. So in your example of increasing local production by 10% the total value would go from 50 to 53 (33+20).
Now for simplicity let’s pretend trade is only split 3 ways, Denmark, Sweden and Prussia. If you as Sweden have 50% of the trade power, and are collecting, Prussia and Denmark are both forwarding that means you would get 25 gold, and 25 gold is getting moved to Lubeck. If Prussia decides to collect at the Baltic Sea instead, with 30% of trade power, then you’d collect 25, Prussia would collect 15, and the remaining 10 gets forwarded to Lubeck by Denmark.
In your theoretical 10% local boost the numbers would be you collect 26.5, Prussia collects 15.9 and Denmark forwards 10.6.
Keep in mind the numbers are rough, forwarded trade does get some value added the farther it moves, so it’d be more like 11.7, but for now we ignore that.
In a trade node with multiple outlets, like the North Sea, trade still splits according to the same rules. So if England has 30 power at North Sea steering to London, Scotland has 10 doing the same, The Dutch have 30 forwarding to Antwerp, and Burgundy has 30 doing the same, then 40% of trade there goes to London and the remaining 60% to Antwerp (nobody is collecting there).
If England and the Netherlands go to war and the Dutch trade fleet is sunk, so now they have 0 power there but everything else is the same, then 42.8% (30/70) goes to Antwerp by Burgundy and 57.2% (30/70 England + 10/70 Scotland) goes to London.
Just remember that for nodes other than your home node your merchants are less efficient at collecting, so unless it’s a very rich node you can project a lot of power at, you’re usually better forwarding trade. Trade efficiency is essentially a multiplier for trade power so higher efficiency means you need less resources to get the same result. Overextension severely hampers your trade collection (to the point that at 100% or higher you get essentially nothing in trade outside your home node).