chore: temporarily bench test deepCopy

[BobbieGoede]

Not sure if this is a realistic test, but it looks like the deepCopy change is slower and uses a bit more memory but I think the performance is still very acceptable.

deepcopy legacy mode false - complex-nested pattern on 1002 resources (AOT):

creation and merging x 1,081 ops/sec ±1.07% (94 runs sampled)
memory usage: rss: 156,332.03 kB, heapTotal: 98,840.58 kB, heapUsed: 37,265.24 kB, external: 2,808.71 kB, arrayBuffers: 41.14 kB

deepcopy legacy mode true - complex-nested pattern on 1002 resources (AOT):

creation and merging x 1,376 ops/sec ±2.30% (92 runs sampled)
memory usage: rss: 131,940.35 kB, heapTotal: 73,674.75 kB, heapUsed: 25,044.62 kB, external: 2,808.71 kB, arrayBuffers: 41.14 kB

deepcopy legacy mode false - complex-nested pattern on 1002 resources (JIT):

creation and merging x 1,092 ops/sec ±0.65% (94 runs sampled)
memory usage: rss: 159,334.40 kB, heapTotal: 101,462.02 kB, heapUsed: 35,273.32 kB, external: 2,808.71 kB, arrayBuffers: 41.14 kB

deepcopy legacy mode true - complex-nested pattern on 1002 resources (JIT):

creation and merging x 1,389 ops/sec ±0.41% (94 runs sampled)
memory usage: rss: 132,042.75 kB, heapTotal: 72,888.32 kB, heapUsed: 30,384.16 kB, external: 2,808.71 kB, arrayBuffers: 41.14 kB

The tests were done by merging an existing messages object with a large one (60000) lines.