I can't pretend to know a lot about avoiding the pitfalls of local currency systems, but I do know that two of the hurdles that you have to overcome are the questions:

• What happens when I'm the only person in the system with a useful skill, and everybody else is making scented candles, and I find myself with as many scented candles as I can ever imagine requiring, and my account enormously and uselessly in credit?
• What happens when the whole system collapses while I'm in credit; do I just lose the value of all that work?

It occured to me that adopting Bitcoin as a reserve currency in such a system might be useful in dealing with at least these two disincenctives to participation, and I was surprised to find very little discussion about it (in fact, I expected that as with most of my other brilliant ideas I would find people who had been already doing it for years, and I was the last person on Earth to know about it). There are some interesting comments in this thread, and a proposal to "fork" Bitcoin-like currencies for local use, which I think is technical overkill but interesting. Here are my initial thoughts:

Anything that makes it harder for the regulatory bodies that have screwed up your local economy in the first place to say "Your local currency is really just an IOU for the national currency so all you people now either owe us tax, or have lost your social security benefits" is a good thing.

Bitcoin would be a good medium for trades between local currency systems, in the same way US dollars are used for global commodities like oil. Note to self: think of an example with less of a bad vibe.

You will need strong capital controls; you're defeating the purpose having a local currency if it's freely exchangable for a global currency. So people should be able to cash out to Bitcoin if the system closes, or they move to another geographical area, or they can cash out a certain percentage of their balance once they've exceeded a threshold, but not willy-nilly. Sure you might then get black market exchanges, but I think you are much more likely to find they will trade in the national currency than Bitcoin, so I don't think controlled exchangability to Bitcoin changes this at all, and since some local currencies have avoided instant failure, I have to assume that under at least some conditions this is a managable problem.

You have to set inflation (and lending, if your system has lending) rates appropriately. A little bit of inflation is a disincentive to hoarding, thereby keeping the local economy moving. And since the value of Bitcoin is increasing relative to any major currency you care to name (and is likely to continue to do so for the forseeable future), you can decrease the value of your currency over time relative to the "real" value of the goods and services you purchase, and also relative to your reserve currency, and still give your participants a high degree of confidence that they're not trading away their efforts for a worthless asset. Worst case scenario is they've earned a bit of "interest" when they cash out, in the form of the difference between the decrease in value of the local currency versus the increase in value of Bitcoin. You can't count on the Bitcoin bubble inflating forever, but the evidence is that Bitcoin isn't going to burst altogether - smarter people than me are calling it "the local currency of the Internet" - so use this Bitcoin deflationary period while you can, I say.

A possible objection to the above interventionist monetary policy is that, depending on who you talk to, either fixed exhange rates (1:1 to Bitcoin, say) or an unregulated currency market will deliver better outcomes than rational planning, because Very Serious People say so. In response, I present the real world. Take a look at it for a while. Provided your system is transparent and democratic, I don't see a better solution than thinking very hard about hard problems. Certainly not faith in gold, dollars, Euros or "free" markets.

In the case of a crisis, the system is going to have to be able to both forgive debtors and pay off creditors. I suppose that means holding at least some Bitcoin reserves. Trouble is, that's an upfront cost for something that you're going to need less and less over time as your local currency becomes more robust. Do new participants "pay" for their initial local currency credit in Bitcoin (a disincentive to participation)? Do established local currency systems either lend their otherwise redundant Bitcoin reserves or go guarantor in order to bootstrap new systems elsewhere? If so, what's in it for them? Do they get to dictate policy to failing systems, like an alternative economy Angela Merkel? Tricky one, that.