Agreed 100%. Statistically speaking, the best approach is to only scan men, as the vast majority of (current) terrorists are male, and the available pool of potential future women recruits is (currently) smaller.
Is it too crazy to expect that sampling for security should match the actual observed distribution (with a uniform prior)?
Is this a bad thing? I mean, romanticize how you will, but is it really better to put human (or even robot) footprints on Mars than to push the telescope programs? 225,000,000 km or 2.25 x 10^20 km - we can't live there (yet).