Almost any time the discussion is about stalling at greater than 1g, it usually involves symmetrical flight, where all portions of the airplane’s structure are experiencing the same g-loading. But what if, say, one wing is at 2g and the other is at 3g, as might be the case in a rolling (banking) pull-up from a dive? The rising wing is experiencing greater g loading because it’s generating more lift. The descending wing, on the other hand, experiences less loading because it’s not generating as much.
That’s asymmetric loading, and the rising wing will stall first, because it will reach its critical AoA earlier than the other one. The structure will survive only if the maximum g experienced by any part of the airframe is less than the load limit. In other words, placing maximum loading on an airframe—by pulling out of a dive, for example—should only be done with the wings level.
If you absolutely, positively must impose asymmetric loading on an airframe, you should not rely on a panel-mounted g-meter, which doesn’t record the maximum load a wing sustains. Instead, it only records loading on the panel-mounted g-meter. A good rule of thumb is to limit asymmetric loading to / of the airframe’s load limit, e.g., use 4g for an aerobatic 6g airplane.