There are a few things at play here.
The VU meters can only get you relatively close. Once the presets are used in a live situation there are a lot of things that can affect their perceived levels when compared to the way they seem by yourself, at home.
The 'Fletcher Munson' effect is one. Cutting some highs and boosting some mids helps here.
Certain frequencies being crowded by other band members is another one. Cutting some highs and boosting some mids can help here too.
Generally speaking, clean tones have more dynamics than dirty tones. In a live situation, we probably all play a little harder? Your clean tones will actually play louder, whereas the dirty tones (because of the inherent compression) will mostly only get dirtier, not louder.
The changing dynamics of the band is another one. Generally speaking, clean tones are 'usually' accompanied by the band playing softer. So the clean tones cut through more easily. The dirtier tones are 'usually' accompanied by the band playing harder. So the dirty tones have a harder time cutting through by default. Sometimes the same preset/scene can cut through just fine in one song, or section of a song. Then not seem to cut through in another the song, or section of song. This has caused me to adopt a preset per song methodology. Even though many of them are just copies of the same preset, their levels, EQ, and effects levels have been tweaked to match the conditions of that particular song.
You have to make (mental) notes during live shows, then go back and tweak some levels and or EQ for the presets. It takes several rounds to get them to a good place, but once you do, it's worth it.