The process id/details you have gathered will just tell you pid on which your JVM is running. For debugging high CPU caused by weblogic managed server, you should be checking on the applications deployed on the server.
At time when you see high CPU utilization, take thread dump on the pid and then using top command get the culprit thread id and check what that thread is executing which caused high CPU.
For taking thread dump refer to - http://middlewaresupport.wordpress.com/2013/01/23/thread-dump-what-when-and-how/
For debugging high CPU refer to article - http://middlewaremagic.com/weblogic/?p=4884
It could be that jstack is not present with the version you are using, but no worries you can select other options mentioned in blog [ http://middlewaresupport.wordpress.com/2013/01/23/thread-dump-what-when-and-how/ ] to gather thread dump.
For the "nid" not matching your thread dump "nid", did you verified the pid on which you gathered thread dump and top command output. If you are not sure about the process id which is running weblogic server you can use $JAVA_HOME/bin/jps -v | grep -i "weblogic.Server" to get the id. If server has been started using nodemanager then go to servers/<server>/data/nodemanager and check for the pid file.
Also remember to gather thread dump and top command output simultaneously.
Steps in blog http://middlewaremagic.com/weblogic/?p=4884 are useful, I was able to match the nid for such scenario in past.