The laws of physics suggest an area about the size of Texas is all the bigger you can make a grid. AC is much better than DC for areas of that range, because you can adjust voltages with a cheap transforms. However if you get too large and feed the grid from multiple places, there gets to be a point in the middle where power from one end is driving the grid negative while the other is driving positive.
You can get around this with DC, and grids do interconnect with DC power, but that requires more expensive equipment.
I'm not sure what you mean by this. The Eastern Interconnection is far larger than the Texas Interconnection and it's connected via AC. Are you saying that the Eastern Interconnection has inefficiencies due to its size? Or are you saying that the Eastern Interconnection isn't fully synchronized somehow?
I tried reading this IEEE article about the time the Eastern and Western Interconnections were connected into a synchronous system (1967-1975) to learn more about the difficulties with scaling interconnections. It mentioned frequency oscillations causing voltage variations close to the East-West connection (p11), which sounds like the issue you're describing. If that is the issue though, then Texas isn't the size limit for a grid because the Eastern and Western grids exist and are far larger.
You can get around this with DC, and grids do interconnect with DC power, but that requires more expensive equipment.