Recently you had an all around update on comments. Great! Finally!
Yet it wasn't anything like reporting comments, or giving us a way of ignoring certain people. But at least a step in the right direction. But one thing in particular is disappointing here:
It seems you did the same mistake again as you did with last_activity.

In terms of last_activity:
If a user edits any comment, commented_at of that user would change.
But you'd have to sync ALL comments ever written by that user because you have no chance knowing what or how many comments were edited (or in this context unfortunately "commented"_at).

In terms of comments/recent:
Again, editing a comment isn't well-thought-out. comments/recent wouldn't change the position if a user edits a comment. Given the fact that last_activity changed I'd expect that comments/recent updates as well and puts the updated comment at the top.
But it doesn't, it stays where it was before, it orders by created_at only.
That means with comments/recent I can only sync comments at a specific point in time. It's basically impossible to sync once and update as needed.
If comments were changed (and I'd have to assume constantly that's the case) I'd have no idea how many comments were changed and I'd have to call all 141640 comments again (worst case) to check.

And why is it that replies aren't in this endpoint? Don't they count as comments in this instance? I feel your wording is highly inconsequential on several occasions leading to issues on where the API goes.

Then there's the issue with likes. Liking a comment changes that comment as well indirectly. But I have no way of syncing/checking that either efficiently.

You seem to create one endpoint after the other to do basically the same thing just with different responses but the same issues.
That reminds me of v1. I'm not seeing the intricate cogs of your API from within but I see it from the outside and what I see is you're doing the exact same mistakes again you did back then with v1...

All we need is a simple way of getting either all or a specific user's comment objects (where replies are considered as comments) and an endpoint to check for comment changes with the changed comment's ids returned, not the timestamp of the last change. The latter is a completely useless information.
Whereas a comment is considered to be changed when: (un)liked, replied to, edited, user rating changed (where possible), or even created.
Endpoint: comments/changed
OrderBy: updated_at
Page and Limit as usual
Returns:
[{
trakt_id: n1,
likes: m,
replies: m1,
user_rating: n2 || null,
created_at: xy1,
updated_at: xy2,
parent_id: 0 || n3
}, ... ]
(Being nitpicky as I am you could then support /comments/[:ids]/* endpoints to lessen the requests, that's what I would do in your position at least)

How would you go about syncing comments of a or all users with your own API as of now? Including updates.
I don't see how that's possible without missing some crucial information or hitting your servers with redundant requests, doing twice or more work on any side involved.

It appears sync/history has stopped working. Im using MediaPortal with app(user/pass) authentication. I believe someone that was using similar auth in TraktRater experienced the same problem, switching to oAuth fixed it.

Can you please look into this, here is an example:

Address: http://api.trakt.tv/sync/history

Post: {"shows":[{"ids":{"slug":null,"trakt":null,"imdb":"tt2861424","tmdb":null,"tvdb":275274,"tvrage":null},"title":"Rick and Morty","year":2013,"seasons":[{"episodes":[{"number":2,"watched_at":"2017-09-21T23:18:03Z"}],"number":1}]}]}

Response: {"added":{"movies":0,"episodes":0},"not_found":{"movies":[],"shows":[],"seasons":[],"episodes":[],"people":[]}}

Headers: {Transfer-Encoding: chunked, Connection: close, Cf-Railgun: f027028b93 99.99 0.052076 0030 e6be, Vary: Accept-Encoding, X-Content-Type-Options: nosniff, X-Frame-Options: SAMEORIGIN, X-Request-Id: 5902f257-0095-41fc-ba59-18f1ae8a3658, X-Runtime: 0.048673, X-Xss-Protection: 1; mode=block, CF-RAY: 3a20bde4134319b0-SYD, Cache-Control: max-age=0, private, must-revalidate, Content-Type: application/json; charset=utf-8, Date: Thu, 21 Sep 2017 23:18:07 GMT, ETag: W/"ea36265e9afc2633c26dd555294d3630", Set-Cookie: __cfduid=dade8663306194ac235c2a3d0e6586cc81506035886; expires=Fri, 21-Sep-18 23:18:06 GMT; path=/; domain=.trakt.tv; HttpOnly, Server: cloudflare-nginx}

When a show is aired on multiple TV network, show each airing date on Upcoming schedule in the Dashboard.

I know the data is from thetvdb and their decision is to keep only the first TV network to air the first episode.

But it's currently a problem with the show "Top of the lake", (https://en.wikipedia.org/wiki/Top_of_the_Lake ). Sundance was the first to air the first season and is considered as primary TV network, but for the second season, BBC as already aired the 6 episodes, and trakt shows the episodes as not aired and season premiere for September 11.

I'm sure it's not the only show where TV networks are collaboring and is aired on each TV networks and it would be nice to have airing date and time on each TV network.

Having multiple TV network could also be useful for searching/sorting by TV network