Use StringPool from CommunityToolkit in ToUpperASCIIFast extension

from be0736b (+ cherry-picked ccba93440) to this commit:
- `Database` reported runtime 0.64 s --> 0.68 s
- profiled `ParseCGIRecord` mean runtime 0.190 ms --> 0.179 ms
- `GameDBHelper` 1M lookups reported runtime 1.03 s --> 0.69 s
- managed allocations (w/o the 1M lookups) 131 MiB --> 123 MiB (fewer
strings but more char[]s?)

since `CompactGameInfo` obviously can't store `Span`s, I tried with
`StringSegment`, but it made both lookups *and init* slower, thanks MS

Directory.Packages.props

@@ -1,5 +1,6 @@
 <Project>
 	<ItemGroup>
+		<PackageVersion Include="CommunityToolkit.HighPerformance" Version="8.3.2" />
 		<PackageVersion Include="Cyotek.Drawing.BitmapFont" Version="2.0.4" />
 		<PackageVersion Include="DotNetAnalyzers.DocumentationAnalyzers" Version="1.0.0-beta.59" />
 		<PackageVersion Include="Google.FlatBuffers" Version="23.5.26" /> <!-- last version with .NET Standard 2.0 support -->

src/BizHawk.Common/BizHawk.Common.csproj

@@ -8,6 +8,7 @@
 		<PolySharpUsePublicAccessibilityForGeneratedTypes>true</PolySharpUsePublicAccessibilityForGeneratedTypes>
 	</PropertyGroup>
 	<ItemGroup>
+		<PackageReference Include="CommunityToolkit.HighPerformance" />
 		<PackageReference Include="Microsoft.Bcl.HashCode" />
 		<PackageReference Include="Microsoft.Win32.Registry" />
 		<PackageReference Include="PolySharp" />

src/BizHawk.Common/Extensions/StringExtensions.cs

@@ -1,6 +1,8 @@
 using System.Linq;
 using System.Runtime.CompilerServices;
 
+using CommunityToolkit.HighPerformance.Buffers;
+
 namespace BizHawk.Common.StringExtensions
 {
 	public static class StringExtensions
@@ -199,7 +201,7 @@ public static string ToUpperASCIIFast(this string str)
 					var c = str[i];
 					a[i] = c is >= 'a' and <= 'z' ? unchecked((char) (c & ASCII_UPCASE_MASK)) : c;
 				}
-				return new(a);
+				return StringPool.Shared.GetOrAdd(a);
 			}
 			return str;
 		}