Well, at least you're closer than most people. I'll give you that much. And the reasons you're off aren't entirely your fault (they're prevailing "wisdom" and I use that term loosely).
1) A very minor factor at best, and one that can turn around and bite the customer in the *** just as easily. Select a domain name that fits the company name (since you seem to be based in England, companyname.co.uk is fine). Personally I don't believe this has had any impact for a long time due to domain squatters, scrapers and other spammers.
2) Site structure is important; you want to keep the navigation as "flat" as possible (no more than 2-3 clickthroughs if you can avoid it) to your major pages. Each of the pages in turn becomes a keyword page, and if the copy is written properly (descriptive header and title tag, quality content, etc.) then it will be fine.
You're right about wanting to focus on key products/services for the customer as well. Put the products/services up that are "best sellers" and focus on those.
Make sure your copy is written for humans first. Search engines, particularly Google, actually appreciate it more (again, this flies in the face of conventional wisdom).
3) 301 redirects are correct. Make sure you not only redirect the main page, but all the appropriate subpages to their corresponding subpages as well.
4) "Sort of". A lot of the times, phrases have lots of results and very few potential customers. The word "mesothelioma" comes to mind...a while back, this was a big phrase for spammers and lawyers because the clients that contracted it could cash in bigtime (and die, but that's not important to spammers and lawyers). Now there are approximately 12,600,000 results in big G for the word, and about 6500 reported cases of the disease. Does that seem a bit off to you?
If you're going to use WordTracker or similar services, the key is to take it with a grain of salt. A lot of keywords/phrases are targeted for the wrong reasons, and often there are pretty good niche phrases that are buried within that take some digging. The KEI (Keyword Effectiveness Indicator) is a pretty good rough gauge of niche phrases, but again...take it with a grain of salt, because a number of the searches are done by SEOs themselves, thus skewing the stats.
These services aren't as useful as they once were, and often quality niche phrases escape them.
5) I find 1-2 works best myself...and I don't beat myself to death trying to get them in there, either.
6) 1-2 different phrases for each page. Remember, each page is a potential entry point, not just your home page. The better you build your site, the less dependent you become on your home page as an entry point, and that's a GOOD thing.
7) Cheap SEO trick. If you write your copy well enough, you won't have to do silly things like this. Does this help your users? If it doesn't, and it usually doesn't, don't do it.
8)
http://www.google.com/search?sourcei...n&q=gas+strike
It's pretty apparent that the title tags within the SERPs aren't 68 characters long (one of the results should be mine, by the way). Some of them are short, some of them are long, and some don't even contain the phrase. Again, this is where "writing for humans" comes into play.
9) That's actually a pretty reasonable limit. Again, though, you may want to expand that somewhat if you can write a meta description that will appeal to users when they see your result in the SERPs. That's trial and error, though.
10) If you mean meta keywords, don't obsess over this one. They're not used, for all practical intents and purposes, by major search engines (except possibly to detect spam). Tier 2 engines do make use of them once in a while, but you won't get much if any traffic from them.
11) They're alt attributes, not tags. Don't let Chris Hirst see that.
Make sure they accurately describe the image to the sight-challenged as well. Don't just stuff them full of keywords and phrases.
12) This is one of those "trial and error" things that depends on the site. Some sites work better with short content pages and some work better with long content pages. It just depends, and there really isn't a universal answer.
The 5 times in the footer thing sounds like the same thing as keywords in the footer. See above.
13) Don't just put in "text rich links" for the sake of them. Put them in where they fit and provide a navigational flow. I'm not sure if that's what you meant by "wherever possible", but it should be. Think about how and where you want to guide the user. What do you want them to do? Do you want them to call? Buy? Email? All of the above? Think of it from that standpoint.
14) If you produce an XML site map, a regular one isn't all that necessary. Most actual users won't even touch the thing.
15) I never bother with this. It's only necessary if your site has errors or issues involving indexing that a sitemap can help isolate. If you build it properly, you don't actually need to do this and it really doesn't accomplish much.
I believe it will at some point, when search engines are forced to adopt "opt-in" logic as opposed to the present "opt-out" logic, but it hasn't happened yet.
16) Good call. Don't be afraid to use meta robots tags as well (you may have to for dynamic content, among other reasons).
17) Link partners and link exchanges are just a silly SEO pet trick that, for some reason, most people think works. They're just a waste of time. Would you rather send traffic to someone in exchange for traffic or would you rather get organic traffic without having to do anything for it?
Search engine submission isn't necessary either. All you need is an inbound link to your site on another site's page that is crawled frequently (e.g. a web design discussion board) and you're fine.
The PageRank thing...toolbar PageRank is useless, and it's not something that should ever be focused on. If you think the directory will send you traffic and you don't have to give it a reciprocal link (one of the signs of a quality directory), then go ahead and submit to it. If you have to give a link back, or the directory seems to be full of crap and spam and stuff you wouldn't recommend to your worst enemy, don't bother. Most directories fail miserably when put to that test.
The kind of directory you want to submit to is
WebSavvy. The owner there is one of the biggest spamfighters around, and you can't even THINK about getting into her directory unless your site's clean as a bean. Build your site to her standards, and you'll generally get at least close to a top-notch SEO site as well.
20) cancels out 18 and 19. If a site owner wants to see how well (s)he is doing, a good site stats package will reveal search engine phrases and where traffic is coming from.
The backlinks don't necessarily factor in, either. Often a quality link to your site will have no backlinks at all that can be tracked, and in some cases may be an internal link (e.g. IBM's blog network).
