Normalization and processing options for RNA-seq data (when no cross-technology aggregation occurs )

### Context

Discussions around species-level compendia #508

### Problem or idea

Related to some discussion on #82

If a user is downloading experiments comprised entirely of RNA-seq data, there are a number of different ways that they might like to receive their data:

* transcript-level counts
* transcript-level TPM (transcripts per million)
* gene-level counts
* gene-level TPM
* gene-level count-scale data that takes into account library size and avg tx length (`lengthScaledTPM`)*
* the `tximport` output (`countsFromAbundance = 'no'`) as serialized R object (`.RDS`) that can be used with `DESeq2`

Additionally, users may or may not want their data normalized using quantiles from the microarray reference distribution.*

*We are set up to quantile normalize the `lengthScaledTPM` output from `tximport`. Our thought was that this is a good starting point for building compendia.

### Solution or next step

This is an area where more user research is required, as it will depend on what folks use for downstream analysis. We should also think about _how_ we would design for these options.

### New Issue Checklist

- [x] The title is short and descriptive.
- [x] You have explained the context that led you to write this issue.
- [x] You have reported a problem or idea.
- [x] You have proposed a solution or next step.

No files, tests, or entry points are identified. Review discussions in #508 and #82 and gather the requested user research to define supported RNA-seq outputs and normalization choices; done means an agreed design and implementation scope.

Written by the indexing model from the issue text.