Ok, that makes much more sense now that I see it.
However, I am willing to challenge this one:
For a bus 1 fault during bus 2 maintenance, several lines would stay connected together on bus 3. With a breaker and a half scheme, only pairs of terminals stay connected.
1. While theoretically true, in most bulk systems (345kv and over) line pairs tend to be far more stable then loosing 2/3 of all circuits. Should both buss bars be lost in breaker and a half, the center breakers will keep power flowing east to west, north to south or visa-versa. At least this is the design topology implemented around here- lines from generation one on side, line to load centers on the other with auto transformers and cap banks on each bus bar.
2. At first glance bus bar maintenance appears more difficult. With the center bus for example you have to stay clear of both end busses close by on each side, and even when dealing with the end buss you have live bays running close below them. Failure or maintenance of a photograph requires taking a bay out of service. Where as compared to BAAH, bus maintenance is a breeze in terms of clearance since one side has no parts of any kind, the other isolated via horizontal disconnected, and live parts kept at a good distance (the feeder cross overs are very high up on the stations I've seen).
3. With centralized bus bar protection common to all 3 busses clearing the entire station is possible if one of the differential relays malfunction. Clearing 2 or more busses is also a possibility since zone discrepancies can occur if any of the pentograph switches malfunction, and generally when bus isolator position switches give contradictory readings the relay is set to simply combine any of the two bus zones in question so a fault on one bus will take out the other over loose connection or broken micro switch.
4. During switching of bus bars a short circuit on a pentograph can also clear two bus bars since zones are summed (combined) during the transfer switching process for obvious reasons.
Just my 2 cents in observation. Some might prove me wrong however.