Welcome to the Q&A Forum

If your website have problem with delay in cached, follow the following steps: Check the crawl error and robots.txt if any through GWT Re-fetched the website through GWT with "URL and linked pages submitted to index" option. Try to increase your website visibility through Social bookmarking and social networking promotion You will get sure benefit through above practice.

That's exactly right.  In fact, a couple of years ago many SEO's believed that sub folders would have more SEO value than subdomains, as it would ensure that all of the SEO "strength", as it were, would pass to the root domain. We've moved away from this now as people have tested subdomains and found that the strength passes equally to them as well.  Whether you go with subdomains or sub-folders, either way the SEO value will pass.

if those pages are all in the same directory are in a couple of them you can remove them using the removal from index tool in gwt, in order to make this those pages should return a 404 or being noindexed (which is what you already did).

There's no real way to estimate how long the re-crawl will take, Ben. You can get a bit of an idea by looking at the crawl rate reported in Google Webmaster Tools. Yes, asking for a page fetch then submitting with linked pages for each of the main website sections can help speed up the crawl discovery. In addition, make sure you've submitted a current sitemap and it's getting found correctly (also reported in GWT) You should also do the same in Bing Webmaster Tools. Too many sites forget about optimizing for Bing - even if it's only 20% of Google's traffic, there's no point throwing it away. Lastly, earning some new links to different sections of the site is another great signal. This can often be effectively & quickly done using social media - especially Google+ as it gets crawled very quickly. As far as your other question - yes, once you get the unwanted URLs out of the index, you can add the robots.txt disallow back in to optimise your crawl budget. I would strongly recommend you leave the meta-robots no-index tag in place though as a "belt & suspenders" approach to keep pages linking into those unwanted pages from triggering a re-indexing. It's OK to have both in place as long as the de-indexing has already been accomplished, as we've discussed. Hope that answer your questions? Paul