I think ECT benefits in some way by its stigma. Whereas the effects of other treatments have been diluted by allowing for "mild" major depression (a little oxymoronic), ECT maintains some pretty stringent standards, which has reduced some bias and noise. As for durability of the response, there's some pretty convincing data that maintenance therapy with lithium + venlafaxine/TCA gives longer outcomes. Not sure if those were represented in the RCTs.
Whats the standard of care in the UK leading to that 33%? Are those numbers for unilateral, ultra-brief pulse, or is bilateral, sine-wave still in use? Also, I haven't seen the data on temporal vs. frontal unilateral, but my understanding is that temporal placement is the standard of care in the States.
I don't think there's any willful ignorance with ECT's side effects, especially considering its been around almost 90 years (including the chemico-electric form). If anything, I'd say its the opposite: the memory loss is one of the most dramatized and recognized of any psychiatric intervention. In this regard, I think physicians need to find a happy medium, but we're still on the overcautious end of the spectrum.