Implement Webworkers in Datafetching

[TheJeran]

The process of grabbing chunks and converting to Float16 can be a bit slow for very large or poorly chunked data.
Since both Zarr and Netcdf4 support parallel I/O, it should be possible to distribute this process into webworkers.

Right now the main thread calls copyChunk2Array() which writes the chunk values to the main data-array.
I think I can create the main data-array with a SharedArrayBuffer
If I understand this article correctly
https://dev.to/rigalpatel001/high-performance-javascript-simplified-web-workers-sharedarraybuffer-and-atomics-3ig1
I can then call copyChunk2Array() in a webworker while passing the shared buffer to it and the worker will mutate the main array.

This should dramatically speed up datafetching.

[TheJeran]

I shoulud also be able to implement this in the
GetCurrentArray() function. This also grabs chunks and throws them into a typed array

[TheJeran]

Also in the Texturemaker where the float16 data is normalized to uint8.