BBad, AIR 87 didn't start in 1987 any more than JP2048 started in the year 2048. There was a lot pie in the sky thinking, but when the proposals were put to government, that was the end of a 3 or 4 year analysis.
Bear in mind that the entire ADF acquisition process has been fundamentally changed twice in the last 15 years, with the Kinnaird review just the latest version of a difficult and unwieldy system. The big problem with Kiowa/Tiger was convincing government we needed a new recce platform at all.
DMO didn't get it wrong any more than CDG got it wrong. Against the criteria set, the Tiger was the best platform. Many people were more upset that we didn't select Mangusta, and the Cobra lobby was quite vocal too. It's one thing to say "we should have bought Apache", it's quite another to convince cabinet of that fact.