Also, check out this section for a laugh: http://www.typography.com/cloud/how-it-works/#the-delivery-a... . "It constructs an application capable of recognizing requests from different browsers." Apparently a few lines of CSS counts as "an application" these days.
Concerning the service itself: according to their product page, H&FJ requires you to include a remote stylesheet rather than javascript. I would say this is the "minimum viable evil" if you are striving to centralize distribution of your webfonts; while annoying, there are some advantages for the user (in this case, free Akamai).
Edit: I just checked a couple of torrent search engines and, apparently, there are not many up-to-date torrents for H&FJ typefaces. In hindsight, I suppose that the population of individuals who are likely to share files illegally and that of professional/dedicated designers have very little overlap.
http://www.fontshop.com/fonts/webfonts/
I suspect other foundries will follow them eventually, because competition from things like Google Web fonts and younger typographers is going to make it very difficult to justify refusing to allow use on the web without a subscription and reliance on a third-party server.
I much prefer paying once for a font and then using it indefinitely, and I'd prefer to compensate a type designer directly for that, instead of paying a foundry. So to me a model where you buy a copy of the font direct from the designer would be more attractive.
I suspect they'll learn better with time, just as other content creation industries have learned or are learning.
It is artificial to use a subscription model for something that is basically a one-off resource. Compare the web font services with licensing fonts for printing or licensing just about any other stock resource for web development, for example.
It seems particularly odd for a foundry that normally aims at the high end of the market like H&FJ to go down this path. Not to belittle all the hard work that no doubt went into these fonts, but at text sizes, on today's screens, on an average quality web site, many visitors aren't simply going to see much difference between a good free font and the most finely-tuned Gotham or Archer, and they're certainly not going to see much difference between a good commercial font from one of the established services that charge significantly less and anything H&FJ offer. However, the pricing model they've picked here appears to be aiming for mainstream rather than the kind of exclusive/high-end sites where their extra polish might be in its element and necessary to maintain the look of quality that the site wants to show.
No doubt they'll be successful anyway, at least for a while, as plenty of places seem happy to jump on the bandwagon at the moment. But I wonder how long this X-as-a-Service gold rush to the cloud will last for before decision-makers start openly questioning whether all these mechanics are just a way of grabbing revenue and whether the extra costs they pay are really justified. Most parts of the tech industry already seem to be going the other way.
For example: The ability to specify a font subset, so that only the 7 glyphs you actually use in your company name can be downloaded instead of the entire font, changing the data moved for just that single transaction from several megabytes to a few kilobytes. Extrapolate that to a handful of faces and a few hundred thousand visits and it adds up pretty quick.
Not to mention you can now stop using Comic Sans. :o)
edit: ok, I didn't read their EULA. Nevermind.