So the mystery was because the original paper had flaws in it's methodology? That's anticlimactic.
Also, not just flaws in methodology, but also flaws in its model, which is a more significant issue. Making better models & checking them against reality is a big part of science. It's not necessarily that the original author fucked up; it's that our knowledge of the world improves (we hope) over time.
Mystery #2! Double climax!
"First, they say, the estimates of 20th-century sea-level rise used by Munk were too high."