His claim is that sin and cos in calculus are often introduced independently of geometry, as math for engineers is non rigorous, and the connection to triangles is a magical coincidence.

See also "Early vs Late Transcendentals" in calculus pedagogy.