I have used Trac for a long time and really appreciate it. I have also looked under the covers and found the sophistication of the coding a little above my pay grade.
If you can dream up a plan which carves out some "easy" bits for people like me I would be happy to volunteer.
Getting a plan together isn't easy. Can I suggest a plan for a plan?
1. Re-document the code in stages with a target audience of ordinary Python devs who might volunteer to port it to Py3
This would be the task of the best and brightest of the existing developers because it would lead to at least a little refactoring. So stage 1 would be top level module doc strings which overviewed each module and explained why it contains the classes and methods it does and its relationship to other modules. Stage 2 would be class doc strings which expand on their module doc strings.
2. Based on re-documentation experience, propose a porting plan among existing Trac devs. Decide whether to use six or go cold turkey.
3. Call widely on existing Trac users to volunteer for the heavy lifting such as unit tests and recoding.
Trac is fantastic software. There are thousands of willing helpers out here.