Currently, all metadata for try jobs are only captured upon completion. If there are errors, much information is lost making determining the root cause difficult. Additionally, we should add whether or not compile targets and heuristic results were used, which are available as soon as the try job is triggered.

1. WfTryJobData should be created as soon as the try job request is made to buildbucket. By then, request_time, has_compile_targets, and has_heuristic_results should all be avilable

2. Try job start_time should be stored as soon as the first response from buildbucket is received, which should contain start_time and try_job_url

3. Upon completion, all other information should become available