It seems to me this is really being discussed as a diagnostic tool. I.e. doing the movement slowly supposedly gives you a greater opportunity to become aware of exactly what is happening at every stage of the movement. The caveat is: is your attempt to perform the motion slowly actually representative of what’s happening when you do it “fast”? Even if it is representative, what is your basis for making corrections to what you discover in the “slow” form? Is there a set of heuristics from a source you trust that prescribe what “good” form is supposed to look like? Where a movement includes a significant ballistic component, is it even really possible to create a “slow” version that’s truly representative of the muscle and nerve activity involved in the fast movement? There are some rudimentary elements in guitar where the answer is probably yes, like “fret the notes with the tips of arched fingers” and “fret right up against the fretwire rather than midway between frets”, but I think for the picking hand it’s much less clear, especially since there are several different approaches that have been shown to work. At bare minimum, I think you need to have a well-defined choice of what the “correct” technique you’re striving for is actually supposed to look like before something like this will be beneficial.
And like a lot of folks, I think tension gets demonized beyond reason in guitar circles a lot of the time. I agree that there are excesses to be avoided, but any time muscles have to do work, there will necessarily be some sort of tension. I think a lot of people use a very narrow definition of tension that entails only that type of tension which is excessive and harmful. But in reality, in any feat of high performance physical activity, some degree of tension will experienced by the performer, even if they don’t notice it, whether we’re talking about a pole vaulter or a ballet dancer. What’s important is recognizing where tension is being helpful and where it isn’t.