Fix fallback sign() issues in Base implementation

[kimikage]

Since the fallback implementation attempts to generate out-of-range values to avoid branching, it throws an error regardless of whether the value is in-range or not.

julia> sign(0.0N0f8)
ERROR: ArgumentError: N0f8 is an 8-bit type representing 256 values from 0.0 to 1.0; cannot represent -1 # before
0.0N0f8 # after

julia> sign(0.0Q0f7)
ERROR: ArgumentError: Q0f7 is an 8-bit type representing 256 values from -1.0 to 0.992; cannot represent 1 # before
0.0Q0f7 # after

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 97.11%. Comparing base (f979b14) to head (66ba27b).

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #360      +/-   ##
==========================================
- Coverage   97.22%   97.11%   -0.12%     
==========================================
  Files           7        7              
  Lines         792      796       +4     
==========================================
+ Hits          770      773       +3     
- Misses         22       23       +1     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[kimikage]

Considering this issue hasn't been reported for a long time, sign() may be unused actually, though.

[kimikage]

Surprisingly, (::Type{X})(x::X) where {X <: FixedPoint} was being called by the fallback method. (I don't think this is an oversight in the test. It's probably the result of constant folding.)
The shortcut might not actually be necessary. (Untested)

Edit:
As for Fixed and Normed, it appeared that the absence of the method did not affect execution speed.
However, it should probably not be removed, as it serve as a fallback in case users or add-on packages define their own FixedPoint types.