I'll preface this by saying that "the important stuff" is clearly subjective. My background is as a mixed Windows/Linux server admin with mainly Windows clients. If you're an all BSD, Linux, or OS X shop, then this isn't really going to apply to you. Sorry about that, I still think you're important if it makes you feel better!
What is one of the most important things to consider when evaluating a piece of software? Simple: manageability. Sometimes manageability means different things to different organizations, but the general premise is the same. How many man-hours will this require for proper deployment, management, and maintenance and are those man-hours worth the benefit that will be gained from the application being evaluated. One of my biggest gripes is when software vendors don't provide a good management platform for their software. Usually this is limited to small specialized vendors, but occasionally a big name enters this arena and I can never understand why. This brings me to today's topic: Firefox.
We've deployed Firefox using the excellent Frontmotion Firefox MSIs for some time now and have generally been happy. My problem is that these are not official binaries in the sense that they do not come from Mozilla. There are also no official ADM or ADMX Group Policy templates. There are, certainly, third party examples of these, but they are generally out-dated and I expect this to be even more of a problem with the recent introduction of Mozilla's rapid release schedule.
I can understand why Mozilla might not want to take on this task. They came from Linux roots and have aimed to make their browser releases consistent across platforms. This would probably cause a substantial increase in the amount of man-hours that would have to be thrown at the Windows version. The problem is that without this, Firefox will never be a serious contender in the enterprise as an alternative to Internet Explorer. I have no idea whether or not this is a real concern to Mozilla, but I do know plenty of IT managers that strongly dislike Internet Explorer. The problem is that they have a hard time justifying Firefox as an alternative, because there's just zero manageability built in out-of-box.
In looking for a solution to my Firefox woes, I came across an interesting piece of information. Google provides an MSI installer and ADMX templates for Group Policy management of their browser. If Chrome can have these features while having a similar rapid-release schedule, as well as *nix and OS X versions, why can't Firefox? Unfortunately, being an educational institution, we deploy things from time to time (like Firefox) to appease faculty members regardless of manageability and just deal with the pains. Higher Ed is generally much different in terms of things being "locked down" when compared to a normal businesses, but the challenges are very much the same. If we weren't thoroughly committed to Firefox as our second supported desktop browser, I'd be testing Chrome deployment right now.
Get on the ball Mozilla. If you don't do it to take market share away from Internet Explorer, you need to do it to keep Chrome off of your back.