Looking at the syslog log files on the front end computers shows:
item 4 has been resolved by reconfiguring the /diskless/root/etc/monitrc file on h1boot1 to increase the high loading limit to 8, and not warn on memory or cpu usage. To install this I restarted monit on these systems.
Carlos and I are working on item 2 (central syslogger)
Keith has a solution in place at LLO for item 3 (ntp drift file) which we will implement here.