Finding bus strip windows breaks when more than one bus is visible

[James Teh]

The code to find bus strip windows assumes that only one bus is displayed at a time. This was the case when I wrote the code, but apparently, this isn't always the case. A way needs to be found to detect the window for the current bus. For tracks, we use the colour of the first 3 x 3 pixels, but this doesn't seem to change for buses.

[James Teh (jteh)]

There doesn't seem to be any visual indication of the focused bus and I can't find a programmatic way to get it either. r54 changes the code to assume bus windows are in sequential order. Unfortunately, this isn't the case when buses are reordered, but this will have to do for now.