A lot of the "progress" is masked by a "Wirth's law"-style effect related to the requirement for quantum error-correction in order to carry out long quantum computations. Briefly, a quantum memory can never be perfectly accurate, and the limit of known physical implementations is close to 1 error in 1000 "clock cycles". In order to overcome this, error-correction is required, which creates overhead. The long-term outlook is still generally positive.