Intent is a hard thing to prove.
In Stormblood, Sheltron impacted your dps. It gave you back MP (which is damage currency), and could also set you up for a Swipe proc. So players incorporated that into their optimization. The main problem with this design is that it devalued Intervention. I think that if you wanted Shield Swipe to work in this way, you would need to do the following:
Sheltron: Block incoming attacks.
Additional effect: Grants Testudo.
Testudo effect: Action changes from Sheltron to Shield Swipe.
Intervention: Redirect attacks on target to yourself. Block all attacks received in this way.
Additional effect: Grants Testudo.
Testudo effect: Action changes from Intervention to Shield Swipe.
Shield Swipe: Delivers an attack with potency X. Recast 1 min. Ends Testudo.
With this setup, it doesn't matter if you're actively tanking or not, you have opportunities to set-up a counter either way. But there is an expectation that you are actively mitigating for yourself and your co-tank rather than sitting on your cooldowns.
There are two design philosophies that you can take here.
1) Mitigation actions should only be used for the sacred task of reducing incoming damage and staying alive.
2) Mitigation should be fun. It should be part of a tank's damage dealing strategy, and you should want to mitigate more damage to set up more opportunities for counterattacks.
I don't mind which of these you prefer. What I dislike is inconsistency.
'Shield Swipe is no good, mitigation abilities are there to prevent damage and nothing more. Oh, but Vengeance is okay because the damage is oh so very little.'
Right. But if it's so very little, then why is it there? Get rid of it, then. People won't miss it, apparently.
Just be consistent.
