I sincerely doubt this will hold up. Why? Because I'll bet they find that, at such minute scales, the measurement of entropy causes a change in entropy, and in this case, the measurement is transferring entropy from the system to the observer. Alternatively, they may have defined entropy incorrectly as "order" when it comes to "beads". I never fully understood all that thermodynamic hoo-hah, but I always felt that "randomness" was an inaccurate analogy for entropy.