Fixed-Point type deprecation

**Type of issue**: Feature Request

Hi! I'm filing this issue about the deprecation of the Fixed-Point type. I believe this is a specific deprecation that falls inside the following category:

> For users who depend on deprecated features for which there is no obvious replacement, please reach out to the Chisel developers by filing an issue: https://github.com/chipsalliance/chisel3/issues.

I am a researcher working on hardware acceleration of cryptography. We heavily used Chisel and the Fixed-Point type for a recent FPGA accelerator we built called [FPT](https://eprint.iacr.org/2022/1635.pdf). We aim to maintain and further develop this accelerator. We want to be able to move to updated Chisel versions in the future, hence why I am reaching out.

I understand that Fixed-Point is a niche and not-so-utilized feature. Are there any plans to add support for Fixed-Point within the new MLIR FIRRTL compiler? If not, is this a "good first issue" we can tackle? FPT was our first project using Chisel and Scala, so we are not that experienced yet.

Another possibility we looked into is to rebuild the Fixed-Point type fully within Chisel, on top of the SInt and UInt types. What is your opinion on that? Why is support preferred or necessary at the FIRRTL level?

I would appreciate any feedback!

Thanks,

**Is your feature request related to a problem? Please describe.**
Deprecation of Fixed-Point type.

**Describe the solution you'd like**
Feedback on best way to migrate a Fixed-Point design.

**Describe alternatives you've considered**

* Re-integrating the Fixed-Point type in the MLIR FIRRTL compiler
* Building a Fixed-Point type entirely in Chisel on top of SInt/UInt

**Additional context**

**What is the use case for implementing this feature?**
Being able to migrate a Fixed-Point design to future Chisel versions.