Security camera deployment and operational issues in the USA

Some time ago I posted about the 'evils' of highly compressing security camera images in order to maximize the number of days of footage that could be stored. That post came to mind as I read an article in the Washington Times online edition that talked about camers to be installed in WDC and referenced similar programs in Baltimore, Chicago, and San Francisco. The first quote that caught my eye was about statistics collected from the Baltimore system and it said:

"Generally, the State's Attorney's Office has not found them to be a useful tool to prosecutors," office spokeswoman Margaret Burns said. "They're good for circumstantial evidence, but it definitely isn't evidence we find useful to convict somebody of a crime."

That quote was followed by this one:

"We have not used any footage to resolve a violent-crime case," she said.
Miss Burns said police sometimes misidentify suspects because the cameras produce "grainy" and "blurry" images.
"We have had that happen more than once," she said.

"Why?", you might ask. The things that come immediately to mind are 1) too far away, 2) poor lighting, 3) poor lens/camera, and 4) too much compression. I'd be willing to wager money on it being the latter, at least in large part. I say that based on experience and also on noting what wasn't said - no complaints about live monitoring - as well as an inference I made based on the fact that a security camera helped solve one case even though no image was recorded (which to me says that an operator likely saw something important using the camera).

The entire article is interesting and gives some insight into real-life issues with deploying and operating these systems. Many of these issues are addressable and data like this can help point the way to workable solutions.