Honestly, that column didn’t hold up well to theory that was current when it was published. It is an example of Cecil’s desire to often titillate rather than inform. There is nothing wrong with having a little fun, but that column doesn’t really give the reader any good information about what would actually happen if he approached a black hole, to wit for anything smaller than a galactic-mass black hole the tidal forces would tear him molecule from molecule, and then the molecules apart.
As for what has changed, aside from some novel but untested (and, at this time, untestable) theories about quantum gravity (i.e. how gravity interacts on the quantum level), there haven’t really been any fundamental changes on the General Relativity side of the house, and Misner, Thorne, and Wheeler’s Gravitation, first published in 1973, is still in print and considered the authoritative work on gravity from a relativistic perspective. The classical thermodynamics of black holes are relatively simple, although there have been some developments in information theory across a event horizon. However, improvements in computational modeling based on GR have permitted higher fidelity and applying very strange non-linear topologies that result from gravitational singularities with various properties. In particular, the interaction between two or more gravitational singularities–traditionally a very difficult problem to solve analytically–can be modeled and examined.
One poster, Chronos, is doing doctoral research in this field and can no doubt offer a more authoritative answer on this topic.