It’s a neat trick and really works well. Why it works is desiccants and desiccant-like materials will pick up certain amounts of water at various humidities so you can generate monotonic curves of moisture content vs humidity. The graph is desiccants but filament will behave similarly.
These are curves but the shapes don’t really matter. What matters is there’s a particular moisture content at one humidity and another moisture content at another. That’s how humidity in a sealed box or bag of filament can be used to get a handle on moisture content. Not a moisture content value, but there will be a value. When we find the go/no-go humidity, we’ve found the go/no-go humidity for that filament moisture content even not knowing the moisture content.
All that said, different manufacturers and colors could have different go/no-go humidities but for a particular type like PLA or whatever, it’s just a guess but I’d bet they are really close. Like I said somewhere here, it’s looking like drying to 24% RH at 55C is maybe enough for PLA while 20% seems necessary for PETG HF (preliminary data).
