Collecting statistics from online public SPARQL endpoints is hampered by their fair usage policies. These restrictions hinder several critical operations, such as aggregate query processing, portal development, and data summarization. Online sampling enables the collection of statistics while respecting fair usage policies. However, sampling has not yet been integrated into the SPARQL standard. Although integrating sampling into the SPARQL standard appears beneficial, its effectiveness must be demonstrated in a practical semantic web context. This paper investigates whether online sampling can generate summaries useful in cutting-edge SPARQL federation engines. Our experimental studies indicate that sampling allows the creation and maintenance of summaries by exploring less than 20% of datasets