Thanks so much for all that information!
I've created an issue to track this format. This corresponds to extract_reads_from_interleaved_file.py in QIIME 1, so there's a fair chance we'll have this in the next few months as we work towards feature-parity with QIIME 1.