Some of this depends on crawl rate and the number of pages you are aiming to de-index. We ran an exercise in which we aimed to de-index 40k pages from our site. It took two months before we saw at least 50% of the pages de-indexed, while it was 6 months before we even saw 75% of the pages de-indexed. However, these were pages that were part of a subdomain with very little authority and low crawl rates. We performed the same exercise with a smaller set of pages, but much higher crawl rates and higher domain authority, and saw an almost instantaneous (24-48 hour) impact in which they were removed. As the mobile site sounds fairly new, you may just have to be patient on it as it will likely take longer to de-index.

We have noticed that Google is indexing the language/country directory versions of directories we have disallowed in our robots.txt.

For example:

Disallow: /images/ is blocked just fine

However, once you add our /en/uk/ directory in front of it, there are dozens of pages indexed.

The question is: Can I put a wildcard in the middle of the string, ex. /en/*/images/, or do I need to list out every single country for every language in the robots file. Anyone know of any workarounds?

Some of this depends on crawl rate and the number of pages you are aiming to de-index. We ran an exercise in which we aimed to de-index 40k pages from our site. It took two months before we saw at least 50% of the pages de-indexed, while it was 6 months before we even saw 75% of the pages de-indexed. However, these were pages that were part of a subdomain with very little authority and low crawl rates. We performed the same exercise with a smaller set of pages, but much higher crawl rates and higher domain authority, and saw an almost instantaneous (24-48 hour) impact in which they were removed. As the mobile site sounds fairly new, you may just have to be patient on it as it will likely take longer to de-index.