AI allows to make live text colored with gradient only by applying new fill to a text. This knowledge often comes occasionally, when users finally got into Appearance. AI just doesn't says it't wrong, shows no hint, does nothing when people try to apply gradient as they do for everything else.

Meantime, each symbol within text can have their own fill.
And when you apply gradient with a drag to an ungrouped set of path objects, AI applies gradient to all of them, placing each gradient into right position.

Thus I just don't understand, why AI still doesn't allow to apply gradient to each symbol within text the same way? So each one of them will have copy of applied gradient, automatically placed in correct position.
Continuing, when two symbols with same gradient are adjacent, their will share center and edge of this gradient.

Just imagine how cool it would be to select three letters inside text and make them gradient when other letters will be colored differently, while text is still live?