Mathematically yes. But to measure 500 Hz you would have to sample at at least 10x the vibration frequency. Assume a sinewave vibration. At the peak(s) of the sinewave, the acceleration is 0, conversely at the zero crossover the acceleration is at maximum.
So at 500 Hz, the full cycle is 2 ms. Your sensor can only read at about 500 ms. Even if the sensor could "average" the full sine acceleration, the result would be 0.