what would be appreciated is
1. sources for your numbers rather than just random numbers thrown out on the internet
2. not using specific city data, i.e. look at national census-type level data. because if you look at the largest cities in the UK you're going to see effects/changes which are local to those cities, and not nationally applicable. manc/brum are still not representative of the whole country.
You're still cherry picking your statistics, which is how to lie with stats 101 (i did that for a job for 3+ years, i know all the wily tricks dude).
Feel free to keep going, but you'll just be cherry picking more and more un-representative samples until you start using national level census-type data.