Improve the performance of project serialization by ~10x

[crsib]

Originally, wxMemoryBuffer was used while generating binary XML data. wxMemoryBuffer grows only by 1Kb, which results in extremely low performance of project serialization (few seconds on Ryzen 5800x for a 100Mb XML file).

This commit introduces a MemoryStream class tuned for high performance and low memory fragmentation.

This issue is related to #2051 but is definitely not enough to fix it.

(short description of the changes and the motivation to make the changes)

• I signed CLA
• The title of the pull request describes an issue it addresses
• If changes are extensive, then there is a sequence of easily reviewable commits
• Each commit's message describes its purpose and effects
• There are no behavior changes unnecessary for the stated purpose of the PR

Recommended:

• Each commit compiles and runs on my machine without known undesirable changes of behavior

[vsverchinsky]

Great improvement overall, though MemoryStream might seem to be a very generic title for solving quite a specific issue, isn't it? I'd suggest moving it under the ProjectSerializer scope.