While some industry and business experts seem to think Fon will do well, some are not expecting wide adoption. Here's a little breakdown for you:

Looks great:

TJ's Weblog

Cool idea, enough money from Martin's prior work, great advisory board with wide acceptance in the community - it seems the company can do the trick if it it overcomes the risks (security and failure to reach a critical mass).

If it succeeds you'll have the next Skype here.

RConversation:

FON actually got started just a couple months ago with a blog post Martin wrote, brainstorming about how one could create a new kind of distributed wifi network. In early October he started the company. In late October he got Boingboing-ed. Now it's a rapidly-growing company, launching today in France, already launched in Spain and Sweden.

Ethan Zuckerman:

FON works by rewriting the firmware on your wireless router, adding an access layer, a billing layer and, ultimately, bandwidth shaping to the router. Right now, this means it only works on one model of Linksys router - in the future, it's intended to run on any wireless access point that can be "flashed" with Linux. This is the same technical strategy a lot of smart mesh networking folks are exploring, making routers run Linux, then running mesh protocols on top of the OS.

maybe:

Om Malik:

Sounds like a kooky hippe free-love type of scheme? I thought so, but I also take Martin seriously. He explicitly calls it his next start-up. He has started a couple of companies before - Jazz Telecom and Ya.com - and sold them like any good capitalist. This is not an open source project. He is hoping to build a giant business out of this.

Read the comments on Om's blogs for more insight from some of the VoIP media. Many are pessimistic, so you'll get a good all-around view of what people are thinking on Fon.