• For use case of one strip per line:
  
  The generated content will need a lot of manual re-timing (as timing information isn't included in the text file).
  
  Further, any edits will then need to be made in the sequencer, as reloading the text will loose all timing information.
  
  Wouldn't it be more useful to have an importer for subtitle .srt files? (we already have an exporter).
  
  Then the strips can always be deleted and re-imported without loosing timing.

• Scene strips aren't suitable for displaying large blocks of text.
  
  Firstly they can't easily be edited in the text field, secondly, there is a buffer is limited to 512 bytes, users who attempt to load an entire text-blocks are likely to run into this limit.
  
  We could have the strip reference a text block directly, as the current sequencer strip limitations means this feature wont work well.

Marking as requiring changes, since I don't think the feature is acceptable in it's current form.

In D10775#274119, @Campbell Barton (campbellbarton) wrote:

• For use case of one strip per line:
  
  The generated content will need a lot of manual re-timing (as timing information isn't included in the text file).
  
  Further, any edits will then need to be made in the sequencer, as reloading the text will loose all timing information.
  
  Wouldn't it be more useful to have an importer for subtitle .srt files? (we already have an exporter).
  
  Then the strips can always be deleted and re-imported without loosing timing.

• For displaying large blocks of text, we could have the strip reference a text block directly, as having to manually sync the text isn't very practical.

The main purpose of this patch is to control where new lines are inserted in the text strips, because the string entry widget for Text strips doesn't allow that.
If it is at some point prioritized to add text directly into the Sequencer Preview in consistency with the current Text Object in the 3D View, this Text to Strip option should be phased out.

As mentioned above has this patch been made in consistency with the existing Text to 3D Object tool.
This tool doesn't support any of the elements you're requesting either.

It is not an attempt to be a subtitle/srt parser/importer. There is already an advanced add-on for that: https://github.com/doakey3/Subsimport and a less advanced(which properly needs an update): https://github.com/gabrielmontagne/blender-addon-load-subtitles-to-vse

Abandoning a patch after a single reply isn't a good way to respond as my reply is not necessarily the final say, it's just raising issues users might run into.

If users want to include a large block of text, referencing text-blocks directly seems a better solution to me.

Besides this, the use case for creating many strips per line/paragraph isn't so clear, as already stated, the effort to re-time each strip make it a one-off operation which can't be easily updated. Do other applications have this feature?