No, it isn’t risky in the least. The analysis makes no use of proprietary data and illustrates potential problems which arise from choices made in the algorithm’s design in an abstract way, using a different word list from the claimed proprietary one that W3W use. The legal threat in question was made because W3W claimed that Andrew Toponce was linking to a product which used their allegedly proprietary word list in a series of tweets. Please don’t make unfounded fear-mongering comments about academic research – they cumulatively have the same effect as the threats you claim to be drawing attention to.