My guess as to why these things might take longer to get more closely integrated - is the certification process. How long would it take to get approval of a new set of controls, software, interlocks and all the rest? The manufacturers will work on the basis that it ain't broke and if the customer ain't demanding it ...?
Yes, basically the requirement has to be included in the aircraft certification standards in order to get things to happen. And for that to happen there usually needs to be evidence of a number of accidents that could have been avoided.
Now in this case I don't immediately remember any other case of a low-level double engine power loss of this "sneaky" type, on approach with autopilot engaged. Those that come to my mind have happened on take-off (SAS 751, MD-82 ice ingestion, and the recent Citation at Farnborough, Kent) which is an altogether different situation. However, the proposed integrated autopilot/autothrottle might also have helped in the "Gimli Glider" case (which occurred at cruise altitude), had the captain not happened to be an experienced sailplane pilot.
Anyway the key issue, as I see it, is whether this type of fault has a high enough probability to justify new safety measures and their associated costs. The investigation will hopefully shed more light on this.