Well I assume the drift based on atmospheric anomalies doesn't tend toward a total delta of +/- 0 over time - or at least we're not accurately accounting for it

Well I assume the drift based on atmospheric anomalies doesn't tend toward a total delta of +/- 0 over time - or at least we're not accurately accounting for it yet. As such we'd face the same january in summer problem the romans faced right? Slowly midnight will move further and further from the time when you're furthest from the sun.