As long as you're just working on JavaScript code for the front end, wouldn't it make sense to write this in node? That way you can avoid context switching. I'd daresay this would be much simpler in JavaScript as well, to the point that a tutorial isn't really necessary.

Here's a quick take. I'm sure it can be cleaned up further. I've also taken the liberty of fixing the delete before post id bug.

    express = require 'express'
    bodyParser = require 'body-parser'
    
    server = express()
    server.use bodyParser()
    
    books = [
      { id: 1, title: "Microsoft Visual C# 2012", author: "John Sharp" }
      { id: 2, title: "C# 5.0 in a nutshell", author: "Joseph Albahari" }
      { id: 3, title: "C# in Depth, 3rd Edition", author: "Jon Skeet" }
      { id: 4, title: "Pro ASP.NET MVC 5", author: "Adam Freeman" }
    ]
    
    findBook = (req, res, cb) ->
      book = books.filter (b) -> b.id is +req.params.id
      if book.length
        cb book[0]
      else
        res.statusCode = 404
        res.send "No book with ID = #{req.params.id}"
    
    server.get "/api/books", (req, res) ->
      res.send books
    
    server.get "/api/books/:id", (req, res) ->
      findBook req, res, (book) ->
        res.send book
    
    server.post "/api/books", (req, res) ->
      book = req.body
      book.id = 1 + Math.max.apply Math, books.map (b) -> b.id
      books = books.concat [book]
      res.send 200
    
    server.put "/api/books/:id", (req, res) ->
      findBook req, res, (book) ->
        updates = req.body
        ['title', 'author'].forEach (f) -> book[f] = updates[f]
        res.send 200

    server.delete "/api/books/:id", (req, res) ->
      findBook req, res, (book) ->
        books = books.filter (b) -> b.id isnt book.id
        res.send 200
    
    server.listen 3000
    console.log "server listening on localhost:3000"

You can allow non-administrator access using netsh[1] as an administrator. That way you don't need to run VS as admin.

    netsh http add urlacl url=http://+:8080/ user=DOMAIN\username

You don't need attributes based routing in this example. Convention based routing is built into WebAPI.

For the Delete method you can return an HttpStatusCode instead of throwing an exception.

    public IHttpActionResult Delete(int Id)
    {
        var result = (from b in ourbooks
                      where b.Id == Id
                      select b).FirstOrDefault();

        ourbooks.Remove(result);
        return StatusCode(HttpStatusCode.Accepted);
    }

Also check out the OWIN self host tutorial [2].

[1] https://github.com/NancyFx/Nancy/wiki/Hosting-nancy-with-owi...

[2] http://www.asp.net/web-api/overview/hosting-aspnet-web-api/u...

Is the normal way to returning "success" values (throwing an exception?)

    var resp = new HttpResponseMessage(HttpStatusCode.Created);
    throw new HttpResponseException(resp);

Apiary is also a good option ( http://apiary.io).

Does anyone know of a similar guide for Ruby? I'm an ASP.net developer and I'm trying to get some insight into other (perhaps quicker/easier) ways of doing exactly what this guide describes.

That's the first time I've heard that meaning of "self hosted." I was expecting something like a REST server that also acted as a client to another instance of itself.