You create text that is basically impossible to search. Like, for instance, do a Ctrl+F on this page and search for “Bold”. You’ll see the example from OP doesn’t get picked up, because it’s not a B, it’s a 𝗕. And it’s not an o, it’s an 𝗼. And so on. Or how about this? Go on Google and copy-paste this word from OP: “s̵t̵r̵o̵k̵e̵”. Now, stroke isn’t a particularly unusual word, but this thread is just about the only result Google returns. Because it’s not stroke. It’s s̵t̵r̵o̵k̵e̵.
It’s also bad for accessibility. A lot of the time screen readers just won’t know what to do with your bold or italic Unicode text.
And of course this only works for characters for which Unicode actually has these variants. Not a problem with the Latin alphabet, but what about Arabic? Cyrillic? Chinese? Devanagari? Hangul? Not gonna work.
These characters are from the Mathematical Alphanumeric Symbols code block. They’re stylized Greek and Latin letters meant chiefly for use in mathematical contexts. The Unicode standard explicitly advises against using them to fake markup for the reasons outlined above and more. A simple markup language is just about always going to be preferable to faking it with Unicode.