1st problem is that the mercury will highly likely poison the fusion plasma. The blanket is already a hard engineering problem to have it work in a "regular fusion reactor". 2nd problem nobody has said anything about is that you get a mixture of radioactive isotopes out of (a supposedly working) a transmutation reactor. This adds to required enrichment processes: to get 100% mercury-198 (from 10%) and possibly a purification of the freshly radioactive gold.

I'd imagine radioactive gold is pretty much worthless. The Au-197 is metastable, and to my flaky understanding will emit gamma-rays, not to mention the other isotopes generated around Au-197.