[Feature Request / Idea]: Why use hashes?

### Suggestion / Feature Request

I saw in the code, you use
```
if hash_file(left_entry.path())? == hash_file(&maybe_in_right)?
```
This reads and hashes the first file and then reads and hashes the second file.
If you anyhow completely read both files, there is no need to apply a hash. You can just take the two Readers and compare byte-by-byte, e.g. something like https://users.rust-lang.org/t/efficient-way-of-checking-if-two-files-have-the-same-content/74735

Note that this is even faster:
- no computation needed to calculate the hash (Ok, usually the bottleneck is the reading from disc...)
- In case of different files, you can early abort and don't need to read the complete files.

Start by locating the comparison code containing `hash_file(left_entry.path())? == hash_file(&maybe_in_right)?`; the issue names no file or test. Replace the hashing-based equality check with byte-by-byte comparison that can stop at the first difference, then verify existing folder-comparison behavior.

Written by the indexing model from the issue text.