I just launched the ability to create "review" posts in Quill, and added support for displaying them in p3k. p3k adds the additional mf1 classes required for the review to show up in Google results as a "rich snippet", although the Google crawlers haven't made it over to my site yet since launching the change.
In the process of adding both mf2 (h-review) and mf1 (hReview) markup to my review posts, it demonstrated that there are some serious issues with the mf2 parsers and how they handle backcompat mf1 markup. Here's an example of my post being parsed by the PHP, Python, and Node parsers:
The Python parser is the only one even close to correct, although it has added a child h-card based on an mf1 object.
I've opened issues on the corresponding repos to try to get these fixed.
In the mean time, my apologies if these review posts appear super broken in whatever reader you're using!
I've had the Echo for a few months now, and despite being skeptical of voice control in general, have been pleasantly surprised by Alexa!
Realistically, I haven't found all of the promoted features as useful as they try to make them appear. So far, it has been most useful for: