This simply isn't true. For instance, if you narrow a variable from `1 | 2 | 3` to `1 | 2`, then trigger a reactive declaration, the reactive declaration can't understand that you've narrowed the variable, and you lose that information. Same thing happens if you go the opposite direction and broaden a variable. TS assumes that program flow roughly goes from top to bottom; if you break that invariant you lose a lot of handy features.

> you clearly don't know what you are talking about.

huh?