If you want to know more about Chroma Subsampling, the
Wikipedia Article is a decent place to start. It describes the common sub-sampling ratios. One thing to keep in mind is that the the ratios described therein describe how it's done digitally. The way it's done with analog video signals like from a VCR or going to an old school TV is done by limiting the bandwidth of the color signal. The analog signals don't describe pixels, per se, but rather, changes in brightness or color as the electron beam of a TV tube is tracing across the front of the screen. The more bandwidth you have available for your luma or chroma signal, the more changes you can accurately fit into the 50-60
microseconds you have to display a single horizontal line of the picture.
It may indeed be helpful to read the various articles on Wikipedia about
Analog Television,
Composite Video, the
NTSC, and
PAL standards to get a feel for how this was done back in the analog days.
Have a look at
this diagram for the signal structure of a PAL analog video signal. You'll note that the portion of the signal labelled "luminance" is a lot larger than the part labelled "chromas". This means while an analog PAL video signal can accurately contain a lot more changes in brightness than it can changes in color. But, that's okay. Our eyes are not sensitive to changes in color anywhere nearly as much as they are in brightness.
The common wisdom here seems to be to capture VHS tapes at 4:2:2, and reading up, I agree. It might perhaps be a little bit more than the amount of chroma bandwidth that a VHS or S-VHS tape can actually store, but it's better than capturing at 4:2:0, losing some of the color information and having your hands tied when you're trying to do restoration work.