To add to the above comment, the analysis provided used the 95 percentile, which isn't what most players will achieve numbers-wise and definitely not the average.

If buffers are more competitive on higher percentiles, it means that "selfish" jobs are likely better at most if not all percentiles because they don't depend on party comp, skill, or size to provide good damage, unlike a job such as BRD that requires the whole group to perform well. The main element that can hinder a non-buffer is undesirable kill times but that's the case for everyone.

This is especially obvious in content such as Criterion and CODCAR (besides NIN) where the "selfish" jobs easily dominate due to having no constraint to deal their damage.

Since the devs cannot balance the game around 90+ percentiles but more average numbers, balance-wise the buffer jobs should always perform better overall at higher percentiles due to the improved buff exploitation by both themselves and the party as a whole. Otherwise, the "selfish" jobs would be overpowered in virtually every piece of content in the game, which as stated above is mostly the case already with the exception of organized 8-people content.

In any case, this discussion is quite old (remember Abyssos?) and the issue is always the same: the extremely limited and slow amount of balance patches within an expansion's life cycle.