Every cell has a full copy of the current operating plan, not the entire history of all preceding operating plans. Storing the entire commit history of our DNA would be much more space intensive.
Well, poking around the web, it looks like the average bacterial time-between-generations is ~0.1-10 hr, and there are ~0.1-100 mutations/generation. So, if life began 4 billion years ago, I get between 10^11 and 10^16 mutations that need tracking.
The article states that there are 10^9 bases in the human genome.
While that's a pretty significant range, it seems that we could store the entire commit-log in the same amount of DNA that 100 cells normally store.. 10,000,000 cells on the high-end. Which is still only a few milligrams of cells. Impratical of course, but interesting.
