It's more complicated than that. Opentree builds a taxonomy which is a 'consensus' from taxonomies from NCBI, SILVA, Index Fungorum, WoRMS, etc. This taxonomy is then used to scaffold the assembly of a phylogeny from the set of accepted trees. These trees (478 in the current synthesis) are selected from the ~3000 studies that have been contributed to the database. There are groups (for example spiders) where the coverage by available trees is rather sparse or absent. In these cases, the synthetic tree necessarily falls back on taxonomy, but that represents a failure of coverage, not the intent of the project.
I agree that the biggest benefit will be to scientists who have a relatively friendly, github backed, collection of published phylogenetic trees (thus, more like TREEBASE). The synthetic tree will help fill in the gaps where a tree is required that is not covered by any one existing phylogeny.