I actually didn't find a cause for the issue, so I had to make a bit of a hack to "solve" it.
I installed a daemon called "monit" and added the following to /etc/monit/conf.d/wamp.monitrc
check process wamp matching "php5.4.6"
start program "/usr/local/bin/wamp" uid 1000 gid 1000
stop program "/usr/bin/killall php5.4.6"
if totalcpu > 90% for 2 cycles then restart
And set the interval in /etc/monit/monitrc to 15 seconds (set daemon 15). As you can see, monit searches for any processes that match php5.4.6 (I renamed the php binary to that) and checks if the cpu usage exceeds 90% for two cycles = 30 sec, and then just kills and restarts the damn thing.
Obviously this isn't an optimal solution, but since this application isn't critical or very heavily loaded, it has worked just fine. The logs say that the script gets killed two to five times a day on average.
However I've since then created another application that uses the exact same codebase for the wamp server logic, but uses a newer version of php (5.5.6) and a newer version of Ratchet, and a slightly different way of initializing the event loop. I haven't had any problems whatsoever with its CPU usage.
Marko