How to compare this type of data?

I am working on a website in dating niche, and I am trying to select a set of 100 keywords for the site. It will be something like "adult dating <city>", "online dating <city>", "<city> dating", "find singles in <city>", and this will be output to 3 million or more pages of the site, even with the smallest cities in the USA. I will have around 2000 keywords like this, and I can run them through AdWords Keyword Planner for most recent search volume, and also check seo competition with SemRush (this comes out as something like 89.6, 75.3, 98.1, numbers like this, lower numbers mean lower seo competition). I can create sets of keywords to run through these tools with Keyword Expander (software), and Notepad++ (search and replace, when needed).

So, in general, I may have 2000 keywords for 10 or 20 cities in the USA, it can be up to 50 too. Search volume will be different for New York (#1 metro area), and Buffalo (this is #50 actually), so this could be factored in, but I am not sure if this will be easy. Seo competition numbers for the keywords (it could be as many as 100,000 total, 2000 per every city, and 50 cities, lets say), may be factored in too, but it can be omitted, as this can become too hard. So it could be only search volume based.

I don't want to get the data, just on New York, lets say. I would like to get a better view, of what keywords are the best, based on 10, 20, or even 50 biggest cities in the USA. If seo competition numbers could be factored in, this would be a perfect thing. 2000 keywords per city to begin with, get search volume, seo competition, and extrapolate the best template keywords, based on that. This would result in a list of 100 keywords, which would be the starting point for a dating site...