The image you posted a link to doesn't say UK anywhere I can see? Google is returning the x-default from your hreflang tags.

Also, from here I see your .com in both the cache and the SERPs. You don't have an Australia language setting so it seems to be working right.

I'm not sure where you're seeing the issue, though. You have three things highlighted in that pic and they are right, right?

Well, having the canonical can help you with other situations (people taking your content, you decide to do translations later, etc) so I would go with canonicals first as they're a more robust solution. Parameter solutions in SC only affect Google itself (not Bing, not any other search engine that comes along) as well. Canonicals would help all of them at once - so def the better choice if possible.

Yes, you should be implementing canonical tags back to the main product page.

Also, if your c.50 URLs are parameter based (ie. /product?color=red) than you can also deal with the indexation of those in Search Console. Google gives you the option to set the options for each parameter. (You can also deal with parameters in robots.txt but unless you have to, I would do it through Search Console instead.)

To set them, go to the Parameters page.

For more information, see Google's help page.

What do the duplicate content URLs look like? In a lot of ecommerce systems you end up with parameter-based URLs such as:

http://www.example.com/products/women/dresses/green.htm
http://www.example.com/products/women?category=dresses&color=green

According to Google "When Google detects duplicate content, such as the pages in the example above, a Google algorithm groups the duplicate URLs into one cluster and selects what the algorithm thinks is the best URL to represent the cluster (and) tries to consolidate what we know about the URLs in the cluster, such as link popularity, to the one representative URL. However, when Google can't find all the URLs in a cluster or is unable to select the representative URL that you prefer, you can use the URL Parameters tool to give Google information about how to handle URLs containing specific parameters." (see more at Google Support)

If your URLs are parameter based I would suggest looking into handling them at that level in Search Console or (last resort) robots.txt as well. However, I'd start with canonicals and parameters if possible.

Tulane.edu, tulane.net, tulane.org & about 15 other sites link to the non-www version.
Worcester, ETSU & AIC (along with 3-4 others) link to https version

Also:
etsu.edu/crhs/aslp/audiology/resources/default.aspx is 404
lakeerie.edu has https issues of their own so the page is missing

In my full look I saw 145 edu links to your site. Of these, hastingscenter looked like it was removed. That leaves 132 unique EDU links.

Of the 132, 108 are not indexed according to SEOTools.
Of the remaining 24, 9 are Worcester, Tulane or AIC.

Accounting for duplicates I see ODU, Setonhill, HamptonU, Pihma, mmm.edu, stanbridge, ccac, lec.edu and lakeerie.edu that should be in Search Console.

You should see some of the remaining 15 in Search Console though. Not sure why you aren't seeing as many are older links.

Haha I read this question earlier, saw the post come across feedly and knew what I needed to do with it. Just a matter of minutes. You're right though - I would've probably said remove earlier as well. It's a toss up but usually when they clarify, I try to follow. (Sometimes they talk nonsense of course, but you just have to filter that out.)

Google's Gary Illyes basically just answered this on Twitter: https://www.seroundtable.com/google-panda-fix-content-21006.html

"We don't recommend removing content in general for Panda, rather add more highQ stuff"

So rather than spend a lot of time on old work, move forward and improve. If there's terrible stuff, I'd of course remove it. But if it's just not super-high quality, I would do as Gary says in this instance and work on new things.

Truthfully, getting Google to recrawl year or two or five stuff can be tough. If they don't recrawl it you don't even get the benefit until they do, if there were a benefit. Moving forward seems to make more sense to me.

A lot of the links you have from .edu sites are linking to the https version of your site. You link to the http version in your profile. When you're looking in search console, which version are you looking at?

I was able to find the links in both Majestic and Ahrefs but many are https ... they may not show up in your link profile if you've only claimed the non-S versions. (I've seen stranger things in Search Console.)

I doubt Google is ignoring them but it may not be displaying all. They don't always show you everything they have and I always take SC links to be more of a representative example than a full collection.

The things you make me Google ...

I agree - and I checked just to be sure but yes, you should be safe. (On the other hand, ExpertsExchange.com changed their domain name after it was mocked about a billion times. They now have the hyphenated version.)