viability of jtag debugging bb-xm with flyswatter2 and openocd?

an upcoming client of mine wants, during his embedded linux course,
at least an intro to JTAG debugging the beagleboard-xm, and given that
a cheap, initial solution is tin can tools' flyswatter2 (fs2) and
openocd, i pulled my fs2 out of the drawer, checked out the latest
openocd code from its git repo, built it for my fedora 20 system, and
got to work learning how to use it.

  the basics work fine -- i can start openocd, telnet to it, and poke
around, but a little searching online turned up a couple posts by
robert berger suggesting that openocd still has its limitations:

https://groups.google.com/forum/#!msg/beagleboard/FhN5WOUwMHQ/tBXqZgCkzeoJ
http://www.reliableembeddedsystems.com/wiki/index.php?title=OpenOCD

  so before i get too deeply embedded into this, can anyone opine on
the viability of using this combination of tools to do *serious*
debugging on a beagle xm? while an fs2 is an inexpensive solution,
does it have any fundamental weaknesses that i would eventually run
into that would make me need to purchase a professional JTAG debugger
(abatron, etc.)?

  more specifically, can anyone attest personally to using this
combination of tools (fs2, openocd) to debug the kernel on a bb-xm?
thanks.

rday