On a medium-long timescale, it has to be basically even, right? If the amount of carbon captured in the soil were secularly increasing at 10% per century, then we wouldn't have any carbon anywhere besides the soil -- even if it was all released at the last ice age.

If the amount of carbon in the soil is secularly increasing at 0.01% per century, well, that's maybe interesting geologically, but for the purposes of handling climate change in the next century or two, we can treat that as equivalent to "it's steady state."