||"Every TV contains a component called the 'local oscillator', which emits a signal when the television is switched on. It's this signal that the external aerials on our vans pick up."
Is this possible/true?
The TV contains several oscillators.
The most important one is the one in the tuner. Oscillates on the frequency of the carrier of the TV channel that the set is tuned in.
Associated to this one is the subcarrier oscillator, which separates the signal components from the signal acquired from the tuner. This one is set to fixed frequency (in Europe, typically 5.5 or 6.5 MHz) that varies with the norm (PAL, SECAM, NTSC).
In color TVs, there is also the oscillator that generates reference frequency for separation of hue and saturation components of color signal (which are modulated as another subcarrier, this time on the brightness signal - the brightness signal is the same between norms, while the color and audio subcarriers vary, which explains why you can get black/white image from PAL channel on a SECAM TV and vice versa). The encoding used is that one variable (saturation?) is encoded as amplitude of the color carrier, while the other variable (hue?) is the phase shift of the color carrier against the reference oscillator (which explains why earlier cheaper color TVs often operated in "false colors", due to poor stability of the reference oscillator). The reference oscillator is put in phase by the means of the "color burst", placed in TV signal just after each horizontal synchronization pulse.
Then there is the horizontal/vertical synchronization pair. These are operating on fixed frequency (15625 and 50 Hz for PAL/SECAM, ??? and 60 Hz for NTSC), and are synchronized by vertical and horizontal synchronization impulses carried in the videosignal.
In modern (read: all but the most obsolete) TV sets there are also the oscillators providing clocks for various microcontrollers inside the TV and the remote.
In modern TVs there are also switched power supplies. These contain another oscillator; they work by rectifying the AC power, converting it to DC (so the diodes and large capacitors rated for 400V), chopping it to about 10 kHz (frequency varies type to type) (so the transistors on the heatsinks), and feeding it to a ferite core transformer (which is much smaller and cheaper and with less turns of wires than an equivalent "low-frequency" transformer), to be rectified on the output side again.
Every oscillator leaks some emissions. Their frequencies and sometimes phases are more or less characteristical; the frequency and amplitude character of the switching power supply oscillator could suggest the type of the TV (or a device in general). The tuner oscillator emissions reveal the channel tuned to. The subcarrier oscillators emissions reveal the norm of the channel tuned to (which has only an academical meaning, as this is an info that is easier to get by other means). The minute differences in emissions from the clock oscillators reveal the activity of the TV's microcontroller - very weak to intercept. The strongest signals emitted are from the deflection coils; their frequency is fixed and characteristical, their phase can indirectly suggest the channel tuned to. The electron guns give out a noise as well, directly related to the brightness of the individual RGB channels; this together with the synchronizations is important for TEMPEST analysis of PCs, if you are into an espionage; though maybe analysis of optical emissions could be easier in amateur conditions. Another source of emissions, both electromagnetical and optical (IR band) is the remote control.
Electromagnetic emissions can tell *awful* lot about you. There is a reason why some more repressive regimes often issue laws limiting non-governmental people and organizations' access to sensitive radio frequency technology operating outside of "officially sanctioned" "consumer" bands, and why TEMPEST-certified devices are for sale only for licenced subjects. Personally I would feel more comfortable if the technology would be on the streets (which would create a large-scale need for defense technologies, that would then appear on the streets as well). If I will get some free time, I will play with the technology a bit, and possibly leak the schematics to the Net; but until then I will have to get some time and equipment, and detailed high-frequency knowledge that I so painfully miss...
 Disclaimer: Pulling it from memory, not exactly sure about the value.
 Some TV "encoding" schemes just drop the synchronizations, and reconstruct them in the decoder from a signal carried typically in other channel. If you will use the color burst to reconstruct the synchronizations (nothing more than a couple of filters, window discriminators, and PLL oscillators), you can watch those channels without having to pay for the decoder.
 Screw the trivia.
 Because the power grids are interconnected, it is critical that the AC voltage has the same phase all over the power grid, otherwise stray currents and huge losses and eventually wire meltdowns result. This was used in the beginnings of the TV, to derive the vertical synchronization signals from. This explains why the vsync frequencies match the frequencies of power grid of their respective areas (USA/Japan, Europe).