I wouldn't describe it as low-effort / high-value, or at least I'd put the emphasis on density of information. We can embed concepts in a very few words, visual information adds a ton of data but not much information.
I often like to compare the post web 2.0 era (ubiquitous, high bandwidth, highly lubricated UIs) with mailing lists (limited, slow, bare). People think more before chatting on MLs, their messages can be short but mean a lot, or can be long and tell even more. Video feels a bit like the former.. you get more data, it's more pleasing for a while, but it doesn't give much more. Sometimes visual / geometric media add some value (when balanced and tuned to massage viewers mental model) but often its just redundant fat.
So much this. That effect is why I cringe every time I'm presented with an instructional video instead of text -- videos waste a lot of time to give me the information that could usually have been done in a page or two of text.
I often like to compare the post web 2.0 era (ubiquitous, high bandwidth, highly lubricated UIs) with mailing lists (limited, slow, bare). People think more before chatting on MLs, their messages can be short but mean a lot, or can be long and tell even more. Video feels a bit like the former.. you get more data, it's more pleasing for a while, but it doesn't give much more. Sometimes visual / geometric media add some value (when balanced and tuned to massage viewers mental model) but often its just redundant fat.