404 is a good way to tell Google that a page is gone, which will then help it drop from index. Removing it from a site map won't remove it from index.

If you can't 301 to a good page for a user experience (for example, it'd be jarring to click on a SERP for a bike tire and be 301'd to a page about women's bike shirts) then there's not a lot of value in doing so.

We don't even bother populating that section any more. Now, that said, Bing does care about them to some degree, so it depends how much you may care about that search engine.

Matt Cutts has specifically said Google doesn't care about meta keywords, so I find it surprising that it'd be used to penalize.

I agree with Oleg; if most everything is the same, even with multiple pages, users may not bother to tell the difference and will contact the wrong site anyway. Thus that creates more work for your internal staff. If you were in multiple countries, different story (perhaps).

And there's less back end to manage, which can be a huge obstacle to SEO if you don't manage it well.

Great points; we have top nav and it's never proved to be an issue. I'm working on a project where we've migrating from left to top nav and I anticipate issues because its a change for the user experience and we've already trained them on a different nav.

The goal should be the fewer clicks to get to the RIGHT item. (Which is subjective, for sure.) People will do more clicks to get to better data, however we shouldn't put them through any more clicks than needed - so, if you can do it in two, great, but if you need 3-4, then do 3-4.

The one thing I have noticed is that (depending on how big your site is) more clicks to get to deeper content can make it harder for search engines to get in and crawl those pages, which can reduce ranking opportunity.

Based on my conversations with Google, relevance is absolutely something to think about - if your site is all about oranges and you get links from sites/even blogs that are all about cars, it's  flag because it can look spammy.

Unless it's an outlier and is just one (or two) random links from sites that aren't related in content to yours, you risk harming yourself by connecting your site with unrelated content.

Actually, if you Google that (free online software that lets you create product catalog layouts) you'll get a lot of options. And that way you can better review what software or set up may be most compatible or easiest for you to use.

Here's the concept at its core: how can Google crawl redirects and index new pages if it can't crawl those redirects to get to the new pages and process the 301s?

Fix that to fix your problem. The link I shared has a lot of good comments very centered on this general topic.

And, I am intentionally avoiding giving an absolute solution to you because, quite frankly, I don't know enough or am involved at all in your site to feel comfortable doing so. Strategically, I'm happy to share ideas/best practices.

I think I get what you mean and this stuff can get a bit tricky - first and foremost, it can take days/weeks/months to get things unclogged after an issue like this and there's no promise you'll get exactly the same ranking as you had before.

Getting back to your original question, and not to kick you when you are down, however, Google never recommends moving an entire site at once because you don't catch major things like this. Now, to your question, here's answer: "In order for Google to index your new site it has to re-crawl the old site which is redirected there. As each url is accessed, the redirection is found and applied." I think that's what you are trying to get at?

There's more info here that may be worth you reading through:  http://googlewebmastercentral.blogspot.com/2008/04/best-practices-when-moving-your-site.html

Personally, I don't go by score. I go by content and what's best for the user and what's most "matched" or "alike" so it reads naturally. My intention isn't to sound smug, but when writing content, I can tell when it's stilted and the same word is used over and over (both my reading and doing a count of it) vs using variations of it. Also, there's tools that I use to at least get an indication if Google might think something's up (i.e., not diverse enough).

Also, my site isn't about one topic - we have a few core things, but we sell more than 30,000 SKUs so it's easier to be more diverse, even if there are similar variations (arc flash labels, cable labels, etc.) which makes it easier.

I track my site/words in a spreadsheet so I can make sure we're not optimizing the pages for the same words and competing against ourselves.

As for backlinks, that can get a bit trickier because I have less control on how people select to link to our site. For that, I monitor and when I feel it's needed, make suggestions and/or collaborate with other sites. (Keep in mind I have 50,000+ pages.)

It's all of those. Those are all ways you seem too fake and "perfect" and it flags Google. Similar phrases (ex: labels, labeling, labeling printers, identification labels, etc.) are all more natural variations of each other that hit home the essential core message (we're know labels!) in a more natural way than having just the word "label" everywhere.

And, if you are interested, here's a nice blog on it: http://www.seomoz.org/blog/301-redirect-or-relcanonical-which-one-should-you-use

And there's less back end to manage, which can be a huge obstacle to SEO if you don't manage it well.

The goal should be the fewer clicks to get to the RIGHT item. (Which is subjective, for sure.) People will do more clicks to get to better data, however we shouldn't put them through any more clicks than needed - so, if you can do it in two, great, but if you need 3-4, then do 3-4.

The one thing I have noticed is that (depending on how big your site is) more clicks to get to deeper content can make it harder for search engines to get in and crawl those pages, which can reduce ranking opportunity.

Great points; we have top nav and it's never proved to be an issue. I'm working on a project where we've migrating from left to top nav and I anticipate issues because its a change for the user experience and we've already trained them on a different nav.

If they can't track their current site, then you are right, they don't care enough. Good thing they can't get what they want, because if it were that easy to de-list a site without having any back end access to it,  then competitors would be doing it to each other non-stop.

If they refuse to make contact then they get to have everything working against them. If you can make contact, then the action items Matt suggested are the next steps.

Good luck with them as a client