Actually, the odd numbers for exposures and f-stops are simply coming from mathematics.  I had time to kill last week and ran out of reading material and decided to leaf through my flashmeter userguide.  In there, they had a table for equivalent EV values. Now, since I always used my flashmeter in aperture priority, I never bothered figuring out wath the EV mode did, but simply put, the flashmeter can read the light intensity and return an integer EV value. Where EV = AV + TV.  Both AV and TV are integer values.

The relation between AV and your f-stop number is:  f-stop = 2 ^ ( AV / 2 )... So:
     for AV = 0, you get an f-stop of 1,
     for AV = 1, you get an f-stop of 1.414 (square root of 2), shortened by 1.4 on cameras
     for AV = 2, you get an f-stop of 2,
     for AV = 3, you get an f-stop of 2.828 (2.8 on cameras)
and so on and so forth.... recognize the numbers ? 1, 1.4, 2, 2.8, 4, 5.6, 8, 11, etc.

The relation between TV and your exposure time is: exposure time = (1 / 2) ^ TV
     for TV = 0, you get an exposure time of 1 second,
     for TV = 1, you get an exposure time of 1/2 second,
     for TV = 2, you get an exposure time of 1/4 second,
     for TV = 3, you get an exposure time of 1/8 second,
and so on and so forth.... so you get 32, 16, 8, 4, 2, 1, 1/2, 1/4, 1/8, 1/16, 1/32, 1/64, etc.

As whaleman suggested earlier, why camera makers write 30 seconds instead of 32 is probably due to the fact that humans prefer nice round numbers. The same applies to the factions, it is easier to remember 1/8 and 1/16 to 0.125 and 0.063.

But the rule of thumb is that from one stop to the next, you always double the amount of light so the exposure time scale has to work the same increments.

But in the examples above, where did the 3.2 and 3.5 f-stop go ?  They are respectively 1/3 and 2/3 stop increments between f/2 and f/4.

I know, I'm a geek.