A short note on plain convergence of adaptive least-squares finite element methods