So basically what you're saying is that Web Design Group, which is a trusted resource on internet coding since 1999 is wrong.  Here's more detail about entities:

http://www.htmlhelp.com/reference/html40/entities/

The ampersand is the first character in an entity.  Entities are well respected and widely used, at least as long as I've been coding web pages (since about 1997).

Ampersand is used as a delimiter for an entity in standard HTML, so inserting it could lead to a validation error and failure to load the page.  If you absolutely must use it in your URL, use the code: & which won't mess anything up.  It's just text, so there's no reason for Google to penalize it.  Under the concept of topic modeling, Google will recognize & as "and" but usually doesn't pay attention to connectors like that, so it's a non issue.

Include everything. You want Google to spider everything.

Let's take an example:

Two pages which talk about Llama Grooming, one which focuses on Llama Fur Enhancement, and the other that focuses on Llama Fur Lustre.

The content for both pages really is the same, just the SEO is a little different for each.

Your main page: www.llamas.com/lfe.com (Llama Fur Enhancement) has all the copy SEO'd for Llama Fur Enhancement.   Your "alternate" page has the same copy, but optimized for Llama Fur Lustre with a Canonical REL tag back to the original page.  If you omit the second page from the Sitemap, it doesn't get spidered, then what's the point?  Both need to be spidered, but the Canonical REL tag tells it which is the "official" page so you don't get dinged for duplicate content.

Capice?

Jim