Fix #1896: Fix report generation when there is only one fuzzer.

[am009]

Fix #1896

When there is only one fuzzer, the df.groupby(grouping2, group_keys=False).apply(is_unique_crash) is something like:

Name: firsts, dtype: bool
c1 firsts                                   0
fuzzer          benchmark   trial_id
aflplusplus libxml2_xml 58        True

which has only one row, even after reset_index

This change uses .stack().reset_index(drop=True) to fix the issue only when the orignal code does not work