No commercial support makes the software a non-starter in most IT environments, that's the (sad) reality. FreeBSD is a good example of why that is a problem - while technically sound, it is at risk of becoming a fringe OS used only by a couple of enthusiasts and FreeBSD developers themselves. For a project that produces an OS, that is a big issue, since you need users to make sure your OS works on various hardware configurations, or else you will become an "experimental" OS like Plan 9.
With FreeBSD being so mature and old as it is, I think it's a given that the chance that it will continue for many years to come is very high. Sure enough, the risk is higher than going with "mainstream" Linux, but if you have the expertise, especially within the business, then it's a no-brainer if you will clearly benefit from all the great things FreeBSD has to offer. We're running FreeBSD with jails and ZFS etc. on a lot of servers and we're very happy with it. It's not hard for a Linux expert to pick up on. Anyway, that's my opinion.