I like the configuration. Their objective is to maintain as much balance between individual cells as possible.
Through multiple charges and discharges, some cells in a bank will get weaker faster than others. This is normal, and you need an equalization charge once in a while (which basically over-charges them all) to bring the weak ones back up to par. Not all battery types can be equalized, so this extra wiring would help extend the life of, say, valve-regulated lead-acid batteries.
The heavier connections on the ends are arranged in a fashion that allows disconnection of some strings, but the fusing doesn't make sense. There are also unequal lengths of wire, adding a miniscule extra resistance between the right-most battery string over that to the left-most string.
I am not certain that the authors of the diagram gave much thought to what happens after one of the fuses blows. Those cross-links would leave a bunch of batteries still in the circuit, even if one fuse blew at the + of the string. Also: since some fuses are in parallel, and others are in series with the rest, I think some faults could cause a "firecracker" kind of failure where one goes first, the rest pick up the extra balance of load, hence overloading them, and so on... Or, the first fuse in the series will blow first, shutting it all down, so what were the other ones for? The fuses do protect the wires, I'll grant them that.
With a bit of work, the cross-linking could be made to work as intended, but they may need fusing of their own - making this pretty complicated.
I think VolvoFarmer and others have said that it's best not to put too many strings in parallel with each other. The author of the diagram may also be aware of this rule of thumb, too.
You said you have 24 cells of 2V that you will hook up all in series for 48V, so all this is pretty academic to you, isn't it?
It's a bit less academic for me, because I actually have 24 cells, (6 Volt VRLA's) and the cross-linking idea sounds neat. I've been measuring quite a variance in cell voltages and this might help. Off hand, I would connect my cross-links with 12 or 14 gauge wire, but I would put a fuse in each of only about 10 amps or so. Or maybe better, that's an application for a "current limiter", instead.
Thanks for posting that.
I only have one turbine, so my system schematic won't deal with your specific question, but it is posted with the rest of my files. It may be of use to you. Your web browser may re-size the image because it is quite large.
System Schematic (as of January 08)