It's indeed a quite original idea, but even assuming that you could somehow to control the electronics inside some barcode reader, I'm unsure if the mirror's wheel mechanics used for raster the laser beam has enough accuracy to assure a precise step, neither the necessary resolution. In addition, should have to provide an external control to move the apparatus in order to perform the surface scanning.
Note that each surface has it's own reflectance properties, therefore you should make experiments to see how behave. Even assuming that you got successful at this stage, in my oppinion you would have problems with thin wires.
Commercial readers have to cater for the barcode or the scanner being moved so the reflected pattern may be stretched or compressed. They use the relative line widths rather than absolute line widths to build the pattern. It follows that it would be difficult to use one to measure distance across an object unless you knew both the wire and scanner were fixed in position and you could gain direct access to the scanner sensor output. On all scanners I have used, the pattern decoder is internal to the scanner unit and all you get out is a serial stream of ASCII characters.
It would probably be more economical and accurate to start from scratch instead of converting an existing scanner.
Consider a magnifying glass to enlarge the wire's image (or its silhouette), across a fixed size window. Then a single light detector can read brightness level. The larger the wire, the more light is obscured, and the dimmer the level.
You'll need to calibrate wire size versus light level.