refactor: improve `symbolic_evaluate` performance

[AayushSabharwal]

Before:

julia> @be (expr, subs) symbolic_evaluate(_...)
Benchmark: 156 samples with 1 evaluation
min    453.542 μs (2318 allocs: 66.359 KiB)
median 501.958 μs (2318 allocs: 66.359 KiB)
mean   605.138 μs (2318 allocs: 66.359 KiB)
max    2.379 ms (2318 allocs: 66.359 KiB)

After:

julia> @be (expr, subs) symbolic_evaluate(_...)
Benchmark: 330 samples with 1 evaluation
min    269.209 μs (1183 allocs: 28.891 KiB)
median 274.625 μs (1183 allocs: 28.891 KiB)
mean   282.744 μs (1183 allocs: 28.891 KiB)
max    588.542 μs (1183 allocs: 28.891 KiB)

Checklist

• [ ] Appropriate tests were added
• [ ] Any code changes were done in a way that does not break public API
• [ ] All documentation related to code changes were updated
• [ ] The new code follows the contributor guidelines, in particular the SciML Style Guide and COLPRAC.
• [ ] Any new documentation only uses public API

Additional context

Add any other context about the problem here.

[codecov[bot]]

Codecov Report

Attention: Patch coverage is 0% with 13 lines in your changes are missing coverage. Please review.

Project coverage is 33.13%. Comparing base (d4b9089) to head (554a212).

Files	Patch %	Lines
src/trait.jl	0.00%	13 Missing :warning:

Additional details and impacted files

```diff @@ Coverage Diff @@ ## master #57 +/- ## ========================================== - Coverage 34.33% 33.13% -1.20% ========================================== Files 7 7 Lines 332 341 +9 ========================================== - Hits 114 113 -1 - Misses 218 228 +10 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.