ASCII has versions of the whole Roman alphabet. I was surprised recently to learn that Unicode contains other version/s of those same characters. One example is "U+1D5C4: MATHEMATICAL SANS-SERIF SMALL K", or "𝗄".
Can't LaTeX math mode, or MS Word equation editor, or whatever other program just use a sans-serif font if it wants the letters in a mathematical formula to be sans-serif?
These characters exist so that the semantic distinction between them can be encoded in plain text, or where the specific font shape can't be controlled.
The block you mention is only intended for use in mathematical and technical contexts, where the distinction between, say, 𝑑 as a variable vs. d as a differential operator vs. 𝖽 as an object (in category theory) is important. TR #25 gives another example where losing the distinction between ℋ and H can completely change the meaning of an equation. Being able to encode this formatting into the text itself is also important for ISO 31-11.
All of these characters maintain compatibility mappings with their "normal" Latin and Greek counterparts, so the distinction between them should not affect searching and sorting.