The current statistics are so simple to the point of possibly being inaccurate. The information comes partially from one build target (like DPDK version) and then from all executions (such as the number of tests run). Let’s include basic information about each build target (version, pass/fail/block/skip numbers) and then a summary at the end.
We could refine some of the SutNode setters when doing this: add setters for SutNode.remote_dpdk_build_dir (so that we don't call the methods multiple times) and SutNode.dpdk_version (as we're setting self._dpdk_version elsewhere)? also for SutNode.get_build_target_info we should probably combine SutNode.dpdk_version and SutNode.compiler_version with SutNode.get_build_target_info
(First pass draft idea) Should break down statistics by: -cumulative pass/fails -per build target: -per testsuite -per testcase(?) Also should list some dpdk artifact/environment info: YAML/JSON can be used.