But he neglected to mention the error bars on this, which AFAIK are huge at least for BH-BH mergers. Every time we built a new instrument, we saw something new, whether in astrophysics, or nuclear physics, or particle physics. Maybe the BH-BH rate is much higher than expected.