Generally I am not a fan of extreme portfolios. I don't think my willingness to take defensive action is a bet on an extreme outcome with a low probability. Aside from the cliche about betting on the end of the world can only pay off once, I think the returns from various foreign markets over the last 10 years tells us that something akin to "regular" portfolio construction still works.
The ideas like "90% t-bills/go berserk with the rest" are interesting on some level, but not necessarily practical. For whatever one of these black swan funds might cost, someone with enough motivation could probably come up with something on their own for much less money.
While anyone who actually wants to do this should remember Montier's advice that you have to decide which tail you are trying to protect against, I think a do-it-yourselfer would probably want a healthy dose of TIPs exposure. 2008 was rough on the ETFs in this space ... but I don't think the panic from back then is likely to happen again. But even so, I would suggest going shorter in maturity, and now there are a couple of ETFs that offer short-dated TIPS exposure include the iShares Barclays 0-5 TIPS Bond Fund (STIP). The foreign TIP space is also intriguing; the new iShares International Inflation Linked Bond Fund (ITIP) could be interesting.
If not foreign TIPS, then explore some other foreign exposure, either with currency or bond funds or, if you have access, actual foreign bonds. Unless the tail you are hedging against is the dollar imploding, then some mix between USD and forex exposure might make sense -- as for now, the dollar still goes up during times of panic. I would think the Aussie dollar (FXA), Canadian dollar (FXC) or Swiss franc (FXF) would work here.
Commodity exposure no doubt has a place at this table, but it is not clear that gold should be the answer. Some sort of legitimate food shortage and gold might not be the answer; again, I think this depends on what tail is trying to be exploited.
Another space that might be relevant is merger arbitrage. Credit Suisse has a few ETNs out there; my wife owns the Merger Fund (MERFX) in one of her IRAs and there are other products out there. There is a pretty good track record with these for inching higher over time in a very boring fashion, which is probably ideal for this conversation.
For anyone who would consider buying a fund that spends some portion of the AUM repeatedly on out-of-the-money put options (that will often expire worthless), then they might also consider a broad-based inverse index fund. If you look at a large sample of them (again, very broad, not sector funds), you'll see that many of them "work" over longer periods of time and have been "working" for a while now. Of course they may not "work" in the future over some extended period, but the result would not likely be worse than continually buying puts that usually expire worthless.
There are obviously more areas to talk about, but the bigger idea is that there are enough ETPs out there to create your own end-of-the-world portfolio and do so relatively cheaply. The most expensive thing mentioned is MERFX, which has a 1.45% expense ratio.
For what it's worth, I still think a relatively normal equity and fixed income portfolio with the occasional defensive action is the best way to go.