Not a clue (I have a vague memory of us delivering a Blue Vixen per month, there or thereabouts, and the duty rumour was that each one resulted in an rough-order-of-a-million quid invoice), and yes - like fitness, it's far easier to keep it than to regain it.How much did the Blue Vixen development cost by the way? Do we need a constant drumbeat of things being developed?
If you're only making radars for a fleet of fifty-odd aircraft (SHAR FA.2), you have to spread years of development cost across fifty-odd sets. If you're making radars for a fleet of six-hundred-odd aircraft (Typhoon), the individual sets can look a lot cheaper. The drumbeat was that engineers moved from Blue Vixen to ECR90 as the projects moved from design, to trials, to acceptance and maintenance. From ECR90/CAPTOR Tranche 1 (with 1991-standard electronics) to Tranche 2 (with 2000-standard electronics) to Tranche 3 (with 2010-standard electronics). Each increase in processor power allowed new radar modes, new radar performance. See now CAPTOR-E / CAESAR, and a range of AESA radars that scales from 20kg and fits in a UAV, to the 250kg monster that sits in front of Biggles.
It's even the same in AFVs. It may be terribly tempting to say "we want a fleet of tanks that are all the same, because training burden and logistics costs", but an MBT is typically in service for 15-20 years (see Centurion, Chieftain, CR2). The "old days" of overlapping Conqueror / Centurion, then Chieftain Mk.3/5 through 13, then Challenger, then Challenger 2, gave a drumbeat of new tank designs, and significant tank upgrades through continuous improvement. We haven't done a new MBT design since, and that capability has left the UK. Our next tank will be designed abroad.