I believe they may have been referring to the Global Intangible Low-Taxed Income (GILTI) tax, which from what I've read is as bad as the OP describes in their post.
In my testing of this query, I ran it against a time range that included over 40 million purchase lines, and our configuration of Redshift returned the result in ~6 minutes. That was much quicker than our legacy EMR implementation.
Currently, we update our product recommendations nightly. However, the speed up we see here from this reimplementation may allow us to update product recommendations more frequently.
I'm the author of the article. At Monetate, we've chosen our data warehouses to maximize throughput, rather than minimize latency. That's where something like Redshift really shines, it's great a large bulk ingests and running large queries relatively quickly, but awful at running lots of small queries quickly.
On our busiest day last year, we ingested over a quarter billion page views across all of our clients' websites. I'm sure someone has made MySQL scale to that volume, but for us Redshift has been working great for a relatively low price point.
This was a talk prepared for the DataPhilly meetup group. My goal was to provide some simple uses I've encountered at my company of mapreduce beyond just the canonical word count example. For the slides that are just a path, please refer to the github repo, https://github.com/jepatti/mrjob_recipes .