Polarization gets you an average corrosion rate on the sample surface. You will have to use some other technique like ultrasonic reflections (which can detect thickness changes) to measure localized pitting provided you use sono-pens with a small contact size.
Potentiodynamic cyclic polarization give you characteristic pitting corrosion potential - Enp and Ecp. Corrosion rate is probably very small and in potential range used to linear polarization or Tafel pitting corrosion doesn't reach. Standards propose valuation of depth, density and size. But you must remember that corrosion rate is related with deepest pit.
Is potential polarization a good method for measuring localized corrosion? No it isn't, assuming that you're primarily thinking of pitting attack (though you should bear in mind that 'localized corrosion' also includes pitting, crevice corrosion, SCC, corrosion fatigue, under-deposit attack, fretting and filiform corrosion).
If you do polarization scans you can begin to build up a qualitative interpretive understanding of when pitting affects the polarization response to a greater or lesser degree on the metal/environment system you are interested in, but as to getting any kind of quantitative information, it is very difficult indeed.
If you want to measure localized corrosion, probably the best approach is microscopy - a surface count of the number of pits and then a cross-section to determine average and maximum pit depths.
However, if you want to monitor localized corrosion, especially pitting corrosion, electrochemical noise analysis is the way to go. Bear in mind that the noise resistance value gives you the total anodic dissolution value. Then you need to do statistics on the signal responses to be able to allocate that overall loss rate to the background uniform corrosion rate, the average pitting rate, and (most importantly) the maximum pitting rate (because that's what controls how soon you get a hole in your vessel wall or heat exchanger pipe, so is the effective prediction of overall service life). Noise analysis is less well-developed than some of the more traditional techniques, but it is quite a bit more powerful.
As Mr. Cox said, microscopy is best for localized. You can calibrate the fine focus for depth and measure distance...you can also try porosity measurements to get an idea of material loss.