It would be useful to be able to differentiate between a failure of the deptrack step to communicate with the deptrack server from a failure to meet thresholds.  It appears from the code that there are comments that the error should be configurable.  See https://github.com/jenkinsci/dependency-track-plugin/blob/master/src/main/java/org/jenkinsci/plugins/DependencyTrack/DependencyTrackPublisher.java#L200

However, I also find it a little odd that the step is directly setting the build result rather than throwing an exception of some sort.  The build will merrily continue on its way with a failed rather than immediately stopping (unless the pipeline owner has decided to try/catch around the call). 

Most other step throw exceptions when they encounter errors so that the build stops executing.