For the sake of completeness, here is what Shearwater said. This was only after 2 preliminary rounds of email back and forth where their response seemed predicated on an assumption that I had OW-level understanding of deco.
I didn't find that completely satisfactory because it makes it sound (to me) like they were just choosing to ignore the off-gassing between 0.79 and 1.0, because it doesn't pose a risk of DCS.
NOW I understand why it works like it does.
The gf99 is not designed to simply show supersaturation, however, it is designed to show tissue inert supersaturation above ambient, total pressure as the risk of DCS only occurs when tissue tension is above ambient pressure. Therefore, the gradient is defined as 0% at ambient pressure, or approximately 1 ATA at sea level. What you are observing is the inert gas pressure fall further towards equilibrium (approximately 0.79). Current models have shown that tissue pressure is not relevant until above ambient pressure and is not represented by the GF99 value.
I didn't find that completely satisfactory because it makes it sound (to me) like they were just choosing to ignore the off-gassing between 0.79 and 1.0, because it doesn't pose a risk of DCS.
NOW I understand why it works like it does.