I wouldn't recommend using the disavow tool for this. The disavow tool is used to clean up spammy links that were not gained naturally.

A better solution is to use 301 redirects and redirect the 404'd pages to the new pages that work on your website. That way users will land where they should if they click the links, and Google will still give you juice from those links.

Here's a place to get started on how t do that: https://support.google.com/webmasters/answer/93633?hl=en

I'd listen to EGOL and Kemp. Sounds like they know the rural places better than I, and these rural people don't know how to search good yet

Wait... the town and county have the same name? Then there's no issue.

People don't search by county and rarely put in the state when searching a city geo. Your money term is variations of, "cleaning services in Winona". Even though this phrase and others like it don't have a high search volume, experience and years of data tell me this would be the way to go. Unless you wanted to focus on carpet cleaning, which is a different ball game.

To sedate your client, maybe discuss a local SEO play with G+. Then you can define the exact area you'd like to cover, which would include both no problem.

I also do this. The meta description is supposed to have a nice sentence or so that is relevant to the page and makes people click. If the content on your page can't do that, you have a bigger problem than meta descriptions.

I've never tried it, but I don't think that would avoid the temporary dip.

It's usually a few weeks, but there are factors that can make that longer or shorter.

Yes. Unless you've noticed that a couple of those domains have gained traction. Then you can redirect a couple, but not all 500.

This is not a good approach. Instead of 500 different domains, it sounds like those domains should probably just be pages or paragraphs on a single domain. You're effectively splitting your link juice 500 ways by trying to get ranked for 500 different domains.

It's impossible for you to take the time and effort on each domain, so the entire campaign will suffer. Instead focus on a single domain, and build that up to cover the areas you want to.

Google knows all about people who try to game the system by buying a bunch of domains. Not only is this a bad approach, it can leads to Google being very unhappy with your brand.

This would be like any other site transfer. Do page-level redirects, expect a dip in traffic after you make the switch, and then monitor to see how well you recover and then determine what other action you need to take based on that information.

Here's an in-depth post that can help you out: http://www.seomoz.org/blog/web-site-migration-guide-tips-for-seos

Consider it this way:

H tags, in a perfect world, would always descend sequentially as content became more specific. So, the H1 and H2 would be more general than an H3 or H4. Also, those H3s and H4s should be inside of the H2s and H1s. This set up allows for more general "head" keywords in the larger H tags, and more specific information that supports the larger H tags to go in the lower H tags. This means you're target area should be in the largest H tag possible, with all subsequent H tags in the list supporting your main headline with proper content. I hope that makes sense.

So, yes. H2 > H4.