I have a 3070 FE, it can reach 70-75 under load, almost 80 Celsius if overclocked with fans at stock speed (I've seen it reach ~240w of power under load). Generally, like the RTX 3060 TI FE it's a power efficient design, at least compared to my previous GPU (R9 390).
I would check the power usage under load, since high power usage is the main contributor to heat (this applies to both CPUs and GPUs). You could experiment with setting a power limit if it keeps hitting 240w at stock, this can be easily set in software such as EVGA's Precision X1.
According to this review, it should reach
around 220w (this is also the official TDP of the card) under full load:
https://tpucdn.com/review/nvidia-geforce-rtx-3070-founders-edition/images/power-gaming-average.png
You can monitor the current / max power usage of your 3070 in games by running GPU-Z in the background during gameplay.
I would try setting the power limit as low as possible initially (see if this affects the clock speeds at all), but leave the voltage /other settings on auto for stability. You can also set a temp limit, which should force a lower power consumption, if the limit is reached.
As for the reason why, GPU dies can vary in quality a bit (feel free to correct me on this!), some may need slightly higher voltages at stock settings.
If still no dice, try removing the side of your case, just to test if this improves GPU load temps (if case airflow is decent it shouldn't matter much).