Why would that be much more effective? Do the hoses have significant count errors, or is it that the RFID system will let them tell what make/model of tire went by and so infer data such as vehicle type that they do not get from the hose system?
That provides a very valuable database for law enforcement, if you know a car was used during a robbery, poof you know where that car as been, so you are one up on the robbers. Even if the license plate was stolen or the car stolen, you have it from the point the plate was stolen to the present. The ALCU has been trying to get statements about how long this data is kept and how it is accessed.
Today you can recognize faces with machines better than you can with humans. A camera can take pictures and store the face data with no other personally identifiable information, and yet when you suspect someone of something you plug in their face data and "poof" you get all the cameras that have seen them and when. Recent laws about trying to protect this sort of abuse not withstanding. [3]
IMSI catchers and simply Wifi MAC address catchers for the purposes of advertising (or surveillance)[4] can provide GPS + Time + identifiable number logs.
Storage is cheap, 32 bit CPU chips are free, HD cameras are cheap, and software radios can build white-space mesh networks on demand. When you lay a grid of these passive technologies around town, it will become the most powerful tool humans have ever invented for keeping track of, or locating, other humans of interest.
The fun part is that private citzens can play too, anyone can put together a fleet of cameras recognizing faces and license plates with a Raspberry Pi 2. And if you have a HackRF One you can pick up GSM bands and WiFi bands to note the phones, tablets, laptops nearby. Call it $400/unit, $40,000 for 100 units (two at every intersection would cover an area of 7 square blocks. Not something you would do on a lark but certainly within reach of someone who could profit from the information.
[1] http://www.kake.com/home/headlines/Salina-police-using-licen...
[2] http://www.governing.com/news/state/sl-surveillance-cameras-...
[3] http://www.law360.com/articles/742169/shutterfly-can-t-shake...
Some folks at UC Berkeley built an RFID reading device you could put by the side of a (one lane) road (like an on-ramp) and read the tires from it. From a traffic analysis standpoint it would be much more effective than those hoses they put across the road to measure cars going by.
At the point we reach Tools Offering Aggregated Surveillance Things (TOAST), you'll be able to figure out where a tire (and the car it was attached to) was 24/7