A perfect example of how you need better segmentation between what you implement for small sellers, and what you implement for large retailers. The combination of SDI (item specifics) along with new analytics for sellers is really geared towards what is being discussed in that article. For what they are aiming for, that approach will work better for mass retail products. One side effect is they may struggle to increase their retailer base as it is may drive average selling prices down in the struggle for visibility.
Curation (better filtering options in niche categories that is specific to the product) generally works much better than analytics for the type of product that smaller sellers sell. If you want a perfect example of this, try finding and comparing items on amazon that are not your typical mass retail product. You get lost in a sea of shill reviews which promote mediocre products to the top of average customer review results and equally mediocre products in the featured section for those with pockets to fund visibility. Some categories have decent filtering options to guide search results whereas others you get a choice of a long list to sort through or mediocre product being pushed on you. Small ecommerce sites that focus on specific niches or categories generally do well because they do a far better job of curating their products than mass ecommerce sites.
That being said I still think most savvy shoppers will continue to use newest/oldest, price highest/lowest searches. Hopefully ebay doesn't drop the ball with replacing those with an all algorithm approach.
