[Task] Refactor errors to avoid delegating `Display` to the source

## Description
Currently the errors of the IOTA Identity framework do not follow the following recommendation from the Rust error handling working group :
> An error type with a source error should either return that error via source or include that source's error message in its own Display output, but never both.

Not following this recommendation leads to duplicate error messages when printing the error along with its causes (using for instance the popular error handling crate `anyhow` or the upcoming [`Report` type](https://doc.rust-lang.org/std/error/struct.Report.html)). See the [What the error handling project group is working towards blog post](https://blog.rust-lang.org/inside-rust/2021/07/01/What-the-error-handling-project-group-is-working-towards.html) for more details.

Example of somewhere where the recommendation is not followed:
```
#[derive(Debug, thiserror::Error, strum::IntoStaticStr)]
pub enum identity_did::Error {
// ..
#[error("{0}")]
InvalidDID(#[from] crate::did::DIDError),
// ..
}
```

Note that the `#[from]` annotation also means that `Error::source` is delegated to the wrapped error `crate::did::DIDError` whenever the variant `InvalidDID` is matched.

Now this code:
```
use identity_did::{Error, did::DIDError};

fn main() -> anyhow::Result<()> {
let err = Error::InvalidDID(DIDError::Other("DID length exceeds the current maximum"));
Err(err)?;
Ok(())
}
```

produces the following output:
```
Error: DID length exceeds the current maximum

Caused by:
DID length exceeds the current maximum
```

In other words the error message is duplicated.

If we instead had defined the variant as follows:
```
#[error("invalid DID handling")]
InvalidDID(#[from] crate::did::DIDError),
```
the output would then be
```
Error: invalid DID handling

Caused by:
DID length exceeds the current maximum
```
which does not duplicate the message in "Caused by".

## Motivation
This is a Subtask of https://github.com/iotaledger/identity.rs/issues/534. Fixing this issue would unlock the possibility to explore using `anyhow` in the Wasm bindings to conveniently produce better error messages there.