I see occasional restarts of the cgminer processor on the experimental one and still try to figure out why it's doing that but restarting solves the issue anyways:
20130331-073001 cgminer restart
20130401-073501 cgminer restart
20130404-193501 cgminer restart
20130405-131001 cgminer restart
20130406-104501 cgminer restart
20130407-194001 cgminer restart
From other users postings I think no systems have failed due to heat or other issues. A few systems were crippled due to transport dislodging cables. One guy had a flaky power supply, easy to fix. One guy had a bad TP-LINK TL-WR703N but was able to merge the modules to another Avalon he had running.
But once the systems are up and mining they are rock solid. Mostly because they running really cool, the input temp is 26 degrees Celcius and the output is only 44 degrees. Compare that to some of the GPU setups....
(So OK, it's only 7 degrees out here, we'll see what the summer brings once it's 30 degrees and warmer)