Firstly, your first and third monitor stanzas above appear to be the same. I don't know of any specific extractions for tomcat logs that are part of any built-in "pretrained" sourcetypes, with that, and because it is so simple to do, I would create your own sourcetypes for each distinct type of data. This will help you validate that the files are getting parsed for event boundaries and timestamps correctly and make it easier to apply your field extractions and other knowledge objects to the different types of data in the future.
I suggest using the data previewer which you can access from splunk web via "add data". This will let you set and validate event breaks, timestamp extraction etc. for each type of data and create the appropriate props.conf stanza for you. In the past, the log4j sourcetype has worked for similar apps for me, but I prefer to create sourcetypes for every distinct type of data.
