Our examples are limited to Movielens (because it's handy), but it is clearly not the most "typical" implicit feedback data.
We clearly need more examples with

• different datasets
• reproducing recent papers