ESA 9162/4 (tuning fork movements) get through a battery roughly once every 18 months in my experience and they oscillate at a mere 300 Hz. So if you were to times this by over 100 the battery life would presumably be considerably less.
This is
hugely misleading:the accutron relies on vibrating a great big lump of metal using electromagnetism while the quartz uses the piezoelectric effect to vibrate quartz. The power requirements are utterly different and cannot really be compared. The Hz isn't what matters here.
Conceptually, it is important to distinguish between the frequency of the quartz and the frequency of the stepper motor. One is the time base, the other powers the mechanical train for the hands and so on.
As for the 5s21, it did, in fact, tick four times a second, but used an impressively Heath Robinson technique (involving springs) to smooth it out)
I describe how here:
http://forums.watchuseek.com/f9/any-qua ... l#poststop
(post four)
It was a victim of the fact that it was an overcomplicated solution to a problem people didn't care much about that was difficult to maintain.
More to the point, the reason that, until recently, we have not seen 'smooth quartz second hands until very recently (in the new Bulova) is simply because a stepping motor had to be used as it was the only way to accurately represent the signal which had been reduced in base two from 32.768hz down to 1 hz as the stepper motor would 'step' by a predictable amount under the pulse of electricity this signal caused. This single step could then drive the rest of the drive chain.
This is where the battery life issue came in: powering the stepper motor accounts for a large amount of the power drain in a quartz watch. Having the stepper motor tick twice a second or more significantly increases the power consumption. Obviously Bulova have solved this problem with their new watches; I'm really curious as to how, but the details are not out there yet.