In order to test the performance, I wrote a test case where the original code was tested against the currently committed CreateObject code and compared against the proposed CreateObject code.
The tests seem a bit inconclusive, because the result will depend on which of the tests is run first (for a reason that is not clear to me - it's not garbage collection as I tested for that too.)
{F1021118}
```
Old CColor JS::Object construction: 113127
Current CreateObject Color: 123451
New CreateObject Color: 109974
Old Grid JS::Object construction: 765894
Current CreateObject Grid: 571565
New CreateObject Grid: 556627
```
```
Old CColor JS::Object construction: 116960
Current CreateObject Color: 133201
New CreateObject Color: 111334
Old Grid JS::Object construction: 150888
Current CreateObject Grid: 113665
New CreateObject Grid: 114157
```
Results do vary. It's not conclusive whether it's objectively slower, than creating the object by hand, but it's conslusive that it's a performance benefit to change the previous CreateObject calls to the new CreateObject calls.