As long as it looks good in SERP and hosts your target keyword you're good.

Yes, google will cut the title at 512px or around 50-60 characters, but if the main information fits within the limit you're good. Plus google still reads and accounts for the rest of your title. Sometimes, shortened title ca even improve CTR, it can work like a click bait title. Think of something like: Best practices for titles are keeping them under 60 chars, adding your keyword and...

But, of course, there's a common practice and it is there for a reason, in the most cases title that fits completely in SERP works best.

Just badges they received for reviews they left. Every review, upvote and etc give reviewer points (like moz points here).

Points: http://i.imgur.com/JoxRVIr.jpg
Badges: http://i.imgur.com/wSSBquq.jpg

I would create lists. So list item markup. 
Something like "25 Best Jokes About Horses", "17 Funniest Puns","Best knock-knock jokes"
Btw, Buzzfeed ranks on #0 for the last one: http://i.imgur.com/bws4qDe.png And they used a simple list item for markup.

I would leave it.
If there is a link to the old page somewhere, you'll get a 404 instead of 301 if you trash it. I'm talking about external links, I guess you've taken care of the internal ones.

If you delete the page any direct or referral traffic to that page will result in 404 instead of 301.

Google shows you the grouped volume. Which means it takes all variations of your keywords, combines their search volumes and shows you which bracket that volume falls into.

Short version: Keyword Planner is s**t when it comes to checking search volumes.

Ungrouped search volume for your keyword is 140 according to Serpstat and Semrush. http://i.imgur.com/Ke6nVqG.jpg

Moz is using clickstream data to calculate the search volume, which means they have actual data on millions of users and their searches and can calculate the number based on this data.

So, Moz data is way more accurate than what you see in Keyword Planner.

Any organic visibility is always skewed. It's an estimate value that's based on data collected previously but it's still valid for comparison.
If you visibility is lower than that of your competitor it means that they rank higher for more keywords with higher search volumes, well sort of. They can rank first for a single keyword with 100,000 monthly searches and if you rank first for a 1,000 keywords with 10 monthly searches, their visibility is still higher.
All queries, including the branded ones are counted, they may not target your branded queries but they can compensate it with higher positions for more popular keywords and their own branded queries.

Here's me trying to explain how it works:

You both rank for 2 keywords each, one common and one branded for each of you.
Common keyword - 10,000 searches
Your branded query - 3,000 searches.
Their branded query - 1,000 searches.

You rank first for your branded query so 100% of people see it and like 90% are likely to click on it. You get 2.8 Visibility points from a keyword with 3,000 search volume.

They rank first for their query and with the same conditions get 0.9 points because their query gets only 1000 searches.

So far you're in the lead 2.8 vs 0.9 visibility points

Now if they rank first for the common query, everyone will see their website an like 70% will visit. They'll get like 7.5 points.
You rank 8th. A lot of people will see your website but only  around 5% will visit it, so you get 1.8 points.

Their total 8.4
Your total 4.6

Now spread this across thousands of queries and you'll know how it works.

Definitely depends on the topic/audience, like you said. It never hurts to run some tests, and don't take anyone's personal opinion on it. I would say that black check marks are better, but I may be 100% wrong in your case.
Lime green frames improve CTR in some banner ads, despite looking hideous and spammy, same for red/yellow text/background combinations and other questionable things.

Vsibilty score is not good for comparing domains at all.
It's more of a metric that you can use to track a progress of a single website. Your visibility will start growing if you start ranking higher for more keywords, which means you're doing good.
As for comparison – it doesn't work because of difference in user intent for each query.
For example we both sell car parts, but I write a buzzfeed-style article about celebrities an their cars and start ranking for a very popular queries like "kanye west car", my visibility will skyrocket, but that doesn't improve my business or means that my SEO is better than yours.

If you don't want to remove these from index and want them to provide more value try adding more fields and up the requirements for the minimum amount of text in the description.

Moz's user profile page is a good example https://i.imgur.com/C4YR7AE.png
There's not exactly a lot of content, but every field is unique for every user

There are alternatives like Serpstat's "related keywords" and Ahrefs' "also rank for". they determine "LSI" keywords by finding connections in SERPs.
Take top 10 for keyword A and compare it to top 10 for keyword B, if 5 pages from SERP A were found at SERP B it's safe to assume that keywords A and B are related.

You can find some cool keywords with these tools
Example: https://i.imgur.com/wO62vII.png and more https://i.imgur.com/HWiBchu.png

Here's what fitment is for those unfamiliar: https://i.imgur.com/RptDZ1e.png

If we're talking branded query here, Google thinks that it's in French, so it shows results in French first.
Is it an option for you to target English speaking audience with a different name like "Group Raymond"?

There a good article on best practices for pagination here, on Moz: https://moz.com/blog/pagination-best-practices-for-seo-user-experience

I personally think, that 10 jokes per page is not a very good practice (obviously depends on how long the jokes are). 
1. 100 Jokes is too much to swallow in one sitting.
2. People don't like pagination when it comes to entertaining content, it always seems like you just want to grab more clicks out of nothing.

I would break those 100 jokes into 5 separate posts, via keyword research, which will help you find neutral niches for your jokes, regardless of your content. Something like "Best jokes for teenagers" (~600 monthly searches) / best new jokes (~ 400 searches, and who cares if they're not really new).

It's better to have 5 different posts like that one, it's easier to rank for these keywords as they're less competitive than just "best jokes", and you'll have a solid semantic structure, crawlers will see that your website is all bout lists of jokes, which is what you really want to achieve here.

First of all, make sure that language comes first when using hreflang,
us-en / pr-es – bad
en-us / es-pr – good

Here's a hreflang discussion you can address https://moz.com/community/q/correct-hreflang-canonical-implementation-for-multilingual-site

And there also should be a presentation by Fernando Angulo of SEMrush somewhere, I was at a conference where he gave a presentation on this matter, it was very useful, try looking it up.