I saw that signed distance(to:) and advanced(by:) were slow, so I'll add my own implementation:
// current
Test Case '-[NBKDoubleWidthKitBenchmarks. Int256BenchmarksOnStrides testAdvancedBy]' passed (0.018 seconds).
Test Case '-[NBKDoubleWidthKitBenchmarks. Int256BenchmarksOnStrides testDistanceTo]' passed (0.014 seconds).
Test Case '-[NBKDoubleWidthKitBenchmarks.UInt256BenchmarksOnStrides testAdvancedBy]' passed (0.003 seconds).
Test Case '-[NBKDoubleWidthKitBenchmarks.UInt256BenchmarksOnStrides testDistanceTo]' passed (0.004 seconds).
// with custom implementation
Test Case '-[NBKDoubleWidthKitBenchmarks. Int256BenchmarksOnStrides testAdvancedBy]' passed (0.004 seconds).
Test Case '-[NBKDoubleWidthKitBenchmarks. Int256BenchmarksOnStrides testDistanceTo]' passed (0.004 seconds).
Test Case '-[NBKDoubleWidthKitBenchmarks.UInt256BenchmarksOnStrides testAdvancedBy]' passed (0.003 seconds).
Test Case '-[NBKDoubleWidthKitBenchmarks.UInt256BenchmarksOnStrides testDistanceTo]' passed (0.004 seconds).
From https://github.com/oscbyspro/Numberick/issues/36