It was looking promising back in 2007, but since it's release in 2008, we knew the promise was empty. The version released was so badly written that even trying to setup a simple product catalog was not possible without hacking the core to fix a few typo mistakes. It was obvious that the system was never used in a production environment, it that sense our own open source eCommerce system Onxshop was already running as a stable software for a couple of years. We didn't have any reason switching to Magento. Yes, it was a version 1.0 in 2008, but reading the comments below nothing has changed in the last 3 years.

Even if Magento was well written, we would still have a big problem that it can only do a product catalog, shopping cart and nothing more. Our clients need usually more than this, they need a complete website, a blog, an user forums, a CMS system. We don't want to add Wordpress for blog, than add Vanilla for user forum, etc. It's very bad for long term support to have a few isolated software packages running on one site. We need everything to interact together. Products, events, news feed, blog, newsletter subscription, user logins and the rest cannot be scattered throughout several isolated systems. E.g. using the same taxonomy system is impossible when running separate systems for one website.

Anyway, Magento future is sealed when bought by eBay. Another hype is over.

Read other people's opinion on why Magento is not what it looks like:

http://igorrac.blogspot.com/2011/07/why-magento-sucks.html

http://www.nicollet.net/2010/06/five-bad-reasons-for-using-magento/

http://php.opensourcecms.com/scripts/details.php?scriptid=92&name=Magento

http://amplicate.com/hate/magento

http://www.pickledshark.com/magento-ecommerce-complicated-bloated-brilliant/

http://www.web-design-talk.co.uk/440/magento-ecommerce-case-study/

or just google "why not magento".