These guys did a killer demo at NYTM. Very cool demo of turning a plant into an instrument (they played an orchid on stage). There was a good question about detection of differing touches of the doorknob, this could totally replace the key some day!
Some of that may depend on precision - how precise is the sensor? Can it pull a fingerprint, so it will basically do a fingerprint scan to authorize door unlocking?
I'm reminded of #66 on the Evil Overlord list - My security keypad will actually be a fingerprint scanner. Anyone who watches someone press a sequence of buttons or dusts the pad for fingerprints then subsequently tries to enter by repeating that sequence will trigger the alarm system.
The authors of Touché did a follow-up paper at UIST this year to study this exact question [1]. In brief, you can train a classifier on the impedance profile of specific users, but there are some significant limitations (impedance profiles naturally drift over time, and accuracy doesn't look so good with a lot of users).
A persons 'capacitive footprint' depends on skin moisture, clothing, anatomy, etc.
This allows Touché to distinguish certain users from each other in a lab setting.
As this approach uses only a single electrode (which is why it is so cool), the data it offers is quite limited compared to more complex sensor layouts.
In real life, such a sensor would need to work reliably even if I wear different shoes, have dirty hands, etc. Therefore, its false rejection rate needs to be very low. Which leads to a high false acceptance rate, i.e. several other people with similar capacitive footprints may open my door, too.
