I've spent enough time on frozen surfaces to know that ice is deceptively heavy and deceptively temporary. Standing on thick ice in winter, you'd think it's permanent - solid as rock - but it's not. Mont Blanc's situation is exactly that scaled up: the summit sits under a dome of glacial ice that's been there for millennia, and as temperatures climb, that ice cap is shrinking year after year. When you're measuring a peak, you're measuring from the highest point, and if that highest point used to be ice and is now exposed bedrock a meter lower, the official height drops. It's not the mountain sinking or the rock eroding - it's the frozen armor melting away.
Here's the practical angle people aren't talking about: the rate of measurement matters. Surveyors can only accurately measure the peak during specific weather windows, and the exact snowline and ice coverage vary wildly season to season. That meter loss in one year might not be purely ice melt - some of it could be that this year's measurement caught the peak in a leaner state (less fresh snow) compared to last year's measurement. The real trend is the decades-long decline, which is undeniable and climate-driven, but year-to-year swings are also partly about *when* and *how* you measure.
If you want a concrete way to think about it: imagine a cake with thick frosting. The frosting is melting. The cake underneath hasn't changed, but the top of your dessert keeps getting lower because the frosting layer is vanishing. That's Mont Blanc.