Welcome to the Q&A Forum

assuming you have the bandwidth i would say host your own and have full controll, both technicaly and for the reasons Dejan presented. i also think that using youtube looks a bit tacky.

Let then 404 and eventualy drop from the index, or remove them from WMT, unless they have external links, there is no use 301'ing them

If you are logged in, google can know if your site is verified under your account in GWMT and know your the owner, but site links is on the search terms used, and the importance of your site.

Take a imgainary site, bobs shoes

If you search for "shoes", you want get such results, you will get an array of sites seelingt shoes with no one getting site links, but if you search for "Bobs shoes" you may, that is because they know you are looking for that company, and are giving you more choices of where you want to go within that companies site, rather then giveing you a choice of sites

i think it looks spammy myself. Just my opinion

I fyou have 301 redirected, then you dontneed to use canonical

In that extreme case, then I would not go near that number, I would not even utter it near a computer or electrical device.
But in general, you can share you ip no with no problems at all. For instance, DiscountASP a big hoster in the US shares almost all of its thousands of websites on one ip number.
Any flagging of IP numbers I believe would be manual, and they would check it out first.
There is something to say about getting all your links from the same ip address, or same c-block of ip numbers, this can look suss if you have too many from the same.

Yes, its like you are saying I want bob's shoe web site no one else, so they will use more of the page to show you more about bobs web site, but simply searching for shoes, why would bobs web site get preference.

Yannick seems to have it right, but here is another way

<system.webserver><rewrite><rules><rule name="DefaultRule" stopprocessing="true"><match url="^index.php"><action type="Redirect" url="/"  =""></action></match></rule></rules></rewrite></system.webserver>

Use canonical tags in all your pages, you are telling the SE that it is your content you have a good chance that search engines will give you the original content credit if they pick yours up first, and if somone steals your content, there is a good change they will take your canonical tag with it pointing back to your site

Without knowing the sites, I would assume it has higher authority than your site. Why does this happen with this site and not others, not all thinks are equal. This is not an unusual event. There could of cause be some problem with your site, where you don’t have the same problem with your other sites.

But the way, for me you come up above trip adviser, when I try, both Bing and Google, US and AU

Maybe i am not understading you, but these pages dont apear to be duplicates top me

I assume you are atlking about the soft 404 errors in GWMT

A soft 404 is a page that returns a 200 when it should return a 404.

Can we get a url?

there is a quict test you can do, see if this returns a 200 or a 404

domain.com/randomText

some websites use the 404 page to do url-rewriting or some other reason, this causes a soft 404.

From what I have read, Google can test if you are doing this buy requesting anonsecne url, and see if it returns a 404 or a 200. if it returns a 200 then they look at your pages more closly.

As many as it takes to write a good piece on the given subject.

Pages do not need a lot of text to rank, Matt Cutts has stated. but having said that, the more text the more chance of long tails

Here is an example
www.lugaluda.com/new-2ne1-i-dont-care-music-video-i-dont-care-lyrics/
has a link to www.lugaluda.com/2009/07/09/new-2ne1-i-dont-care-music-video-i-dont-care-lyrics/

You can get site links for subdomains as well as subfolders. This is not a problem.

Look at this example

https://www.google.com.au/webhp?hl=en#hl=en&output=search&sclient=psy-ab&q=iboats&oq=iboats&aq=f&aqi=g4&aql=&gs_l=hp.3..0l4.1161l2474l0l2774l6l6l0l0l0l0l581l2005l3-4j0j1l5l0.&pbx=1&bav=on.2,or.r_gc.r_pw.r_cp.r_qf.,cf.osb&fp=1a7137b3da639030&biw=1280&bih=881

There is no setting or feature to do this, but you can do it with code
You could detect the useragnet and create a robot.txt dynamically depending on the useragent.
But i cant see why you would want to do this.

I know that many have said that over time, but i dont think it is true, recently GWMT has started to show links comming from subdomains as internal. As long as no one else has verified the sub domain it is assumed to be owned by the root domain owner.
Also site links in google contain sub domain links. so i believe that SE's see them as the same site.

Also Matt Cutts has said there is no difference, only to say that you should have enouth content to warrent it

Thats is good news, it means you have room to improve, you dont want to find you have every thing spot on and still dont rank.

I believe you are right, i think i have a site that is been hit, yet I cant see what couild be wrong.

Maybe i spoke too soon

You mean he has lost his domain name?

He can not 301 a domain he dose not own, he need not worry about duplicate content as the old content will drop out of the index eventualy.

If he still has access to the account in GWMT, he can remove the site, but i would hurry before WMT tries rto validate the site again and finds out it no longer exists