I love parallel processing, but I want to run parallel nodes filtered by specific features of columns, similar to how Group Loop Start can split up a task.
Can we make this happen? 
This would be very useful in speeding up the processing in KNIME Reporting.