Support providing datasets and copies in the upload flow

In the demo app, user always has a dataset already create before uploading, with the multi sp upload flow it doesn't work anymore because **i can't say use this dataset to put my files and i want 2 copies (product requirement)**.
Synapse SDK will return an error saying providerIds array length doesn't match the number of copies.

This goes back to my proposal [here](https://github.com/FilOzone/synapse-sdk/issues/495#issuecomment-3633478315), the current upload options ( datasetIds: array, providerIds: array) makes this type of use case hard to do.

If the user wants to say put this files in these datasets (1, 2, 3, or more) and make sure i have datasets.length + 1 copies the current api does not support this.

/cc @rvagg

Start with the demo app's multi-provider upload flow and the SDK upload options, especially datasetIds and providerIds. Review the proposal in issue #495 and trace how the SDK validates providerIds against the requested copies. Done means the upload flow can target one or more existing datasets while requesting the required number of copies without the array-length error.

Written by the indexing model from the issue text.