[amelius]

Ok but imagine a circular grid of N nodes. Following the circle, the stations are connected to the grid. Of course there is a small, inevitable mismatch at every connection. Then, when connecting the last node, you might get a problem which can only be solved by adjusting the settings at the other nodes.

[pjc50]

It turns out that this can be a substantial problem, resulting in wasted loop flows: https://iopscience.iop.org/article/10.1088/1367-2630/18/10/1...

Generally systems try to avoid too much loop power in the first place for this reason, setting up a directed but acyclic graph with some redundancy links carrying zero or near-zero power as standby.

[rtkwe]

The grid doesn't really pass through each of the points though so you don't really get the frequency mismatch between one side of a node on the grid and another. Instead of a ring of nodes it's closer to a ring with nodes hanging off of it so you never really get two frequencies observed at one power station because it's only connected to the rest of the grid at a single point.