I have some actual experience with this subject as an employee of a company that installs and supports various recording equipment for various public safety organizations - mostly the local, county, and state police and mostly the multi-line phone/radio recording equipment. The company I work for also installs and supports interview room video recording equipment, handheld and remote dictation equipment, and a bunch of other stuff that may or may not be geared towards the public safety sector.
We looked into installing and supporting officer-worn cameras but the market is actually pretty dry from a reseller point of view. There's really only two major players in that space with equipment that agencies actually want to buy (and one of the companies rhymes with "phaser"). All other manufacturers are basically joke products, mostly due to their inability to compete on bulk (or, specifically, lack thereof), price, or features. The two major players aren't talking to resellers right now - one is in-house only with their own nation-wide (and probably world-wide) sales staff, the other is just too busy trying to get resellers signed up across the nation.
Background out of the way, the major problem as I see it is logistics. The political issue of recording everything an officer does while on-duty is pretty much settled at this point - it's going to happen. But how does an agency handle all that raw data? Police forces are typically run on a shoestring budget and rely heavily on donations. The cost of implementing a storage mechanism for all that raw video, with backups and redundancies, for their mandated retention period, is not a trivial task.
And don't think the agencies can simply go out to Amazon, Google, or Dropbox for cloud storage. That video is considered "criminal justice data" and cannot simply be stored on a commercial-grade cloud. There are specific rules on data access such that it must be 100% controlled by the entity that generated it (read: police). Even non-police IT staff from the city, county, or state have to be certified for doing IT work for a police agency since they're exposed to criminal justice data in the course of doing their job. Granted, it's not terribly difficult to get that certification, but the rules about what you can and cannot do and the punishments for violating rules are quite lengthy. Every state in the U.S. has very similar rules since it's ultimately driven by, and approved by, the FBI - all the way up to the director. Long story short, agencies need their own storage system with their own, dedicated off-site backups that only certified people can access - physical, logical, or otherwise.
So the purchase of the camera system is the first major expense. Training on how to wear and use the cameras also incurs a substantial expense. Storage is going to be wicked expensive. Let's then move into how to actually USE the recorded video.
I don't think the body cameras are going to be streaming all the recorded video in real-time to the central repository. Right now, I'm betting all the video is stored on the device in a memory buffer. Once the shift is done, the camera is likely placed into a dock of some sort for offload. Even if a system is developed so the recordings are immediately offloaded or streamed to a car-mounted storage mechanism, it still needs to go back to the central storage repository. That makes real-time analysis of the video quite a bit more difficult. I would think the initial use of the video from the body camera is going to be quite similar to that of the phone/radio recording solution - analysis after-the-fact for training and/or used as evidence in a court case. Both situations are going to occur quite a bit after the video was recorded.
What they're (probably) NOT going to do is mine that video for use with LPR software. As it is, the LPR software is tied to the squad cars so officers in the field are more aware of their surroundings. Feeding a near-constant stream of video from dozens of officers per day just to keep track of a handful more cars and their potential locations at a given point in the day is not exactly a cost-effective (money nor time) way to use the cameras. The video is simply going to be available as another source of evidence when dealing with conflicting reports of what the police or subject(s) did or didn't do.