Are these actually lithium ion batteries? I kind of assume so given the fire. Why not use something like LFP instead, for a grid-scale installation where weight doesn't matter so much? Along the same lines, is the Tesla Powerwall a lithium ion battery? Most DIY equivalents are LFP, that I'm familiar with. This seems like a sensible choice.
Because Tesla doesn't have LiFePO4 (which, technically, is a category of lithium-ion, just a less exciting variant) production, and isn't using them in their cars, so they don't have a bunch around. They're using a low-cobalt variant of NMC, last I checked, though that may have changed to NCA at some point. They're not nearly as exciting as LCO (lithium cobalt oxide, high energy density and really, really happy to come apart at the seams), but still can runaway pretty happily if they get hot enough.
You can get LFP to runaway as well, it's just a lot harder, requires far higher temperatures, and doesn't release nearly as much energy.
LFP is good for grid storage, though depending on what you're doing, if you don't mind a bit of watering (or catalytic recombiners), flooded lead acid is pretty darn boring. Not the most efficient (though if you run it from 30% to 80% SoC like lithium, it's quite efficient, and the newer lead-carbon combinations resist hard sulfation pretty well), but it lacks the exciting failure modes of lithium.
Especially if you have flooded lead acid, the battery just doesn't have the energy to boil all the water out. The only real risk is hydrogen buildup, and that's easy enough to mitigate by giving it "Up and to the outside" path - it goes up really, really well and a bit of fresh air plus a path up mitigates just about all hydrogen buildup issues.
There's no particularly good reason to use a high energy density lithium cell for grid storage unless that's just what you have laying around, which was obviously the case for Tesla.
And, apparently, their cooling system can't handle a runaway as well as they thought.
That's a good question. I suspect in this case it's a matter of Tesla having a large supply of a certain kind of cell at low cost, so that's what they used in Powerwalls. If someone else can make a product that is cheaper, safer, and lasts much longer by using LFP cells, then there won't be any reason to buy Tesla's product. In the long run I expect LFP (or whatever supersedes LFP) to win in the end.
I don't know if there are any LFP-based Powerwall-equivalent products, but I assume they must exist at this point and are probably being deployed in utility power storage projects. I just don't see why you wouldn't use LFP unless you either can't get them for a reasonable cost, or these things are being deployed in unheated facilities that experience below-freezing temperatures (LFP generally can't be charged in sub-freezing temperatures).
