Seriously, why? Plugins are less well-supported and generally much slower to update. They're easier to install, to be sure, but you've already installed it. This really seems like it would be a step backward.
I don't disagree at all, unless the objectives are to achieve an 'appliance' feel through the use of the comprehensive UI and to minimise 'tinkering under the hood' via the shell console. The less techy I can make FreeNAS appear to be, the more likely I'm able to convince a less technical audience to take it up. The techier it appears, the more I'm likely to be tied in to support those deployments. I'd like to loosely couple support of those deployments as much as possible.
On a personal level, I guess I'm of the ilk that I prefer the ‘appliance’ feel of FreeNAS with just the occasional need to roll up my sleeves and look under the hood. No doubt, there will be others that argue the other way (go FreeBSD I say), and everybody else will argue somewhere on the spectrum in between these extremes.
As a for instance, Plex is a storage hog. A less tech-savvy audience will naturally, primarily because of platform familiarity, gravitate towards a Windows or Mac solution for a Plex Media Server (PMS) before they consider a FreeNAS solution. What's not on the radar for them, and not an easy sell as they are somewhat abstract concepts, is the stability (long server uptimes), data integrity and redundancy (based on ZFS), and backup functionality (data replication, snapshots, rollback) that FreeNAS offers. FreeNAS is perfect for a PMS, but not if I'm tightly coupled into supporting those solutions.
Being on a Plex plugin version (currently 1.18.3.2129) compared to the latest version (1.18.6.2368 released today) is not a dealbreaker for me. This will be true (most of the time I hope) for other plugins as well. On the other hand, tying myself into supporting 'under-the-hood' solutions is a dealbreaker for me.