@AntiSC2 Please forgive me if I am wrong here, I do not know whether you had gotten permission or not, however I am not sure whether anyone but the Triage Team (which from what I see, including @Girtablulu) is allowed to add or remove tasks to/from projects.

And an update, the same exact symptoms also occur (again with a 9260) on the Acer CB3-131.

By the way, after reading similar issues being posted over the years at the Arch forums, I am starting to have suspicions that this issue may be related to T4351.

Now that I have gotten a new Fenvi FV-102 M.2 to PCIe WiFi/BT adapter (I had given my previous one to my sister, along with one of my Intel 9260s), I can again test this on my desktop.
I can confirm the network speed issues are persisting on my desktop, using the Intel Wireless-AC 7260.

Alright, I am starting to get results with Qualcomm.
I seem to be getting similar internet slowdowns with my Qualcomm Atheros QCA9377 (using the ath10k driver), however they are not as sharp as with iwlwifi.
Internal network connection speed always shows either 1MB/s (even when it is faster) or "Unknown".

I still need more time to test system stability with this card.

Send a ping to your router should the ping be 10 or higher then something between you and the routher is wrong. Btw i see on some german forums that your problem also occurs on Windows 10

Thanks, my worst ping result is generally ~7.5ms (including my test right now, best ~1.5ms, average ~4.6ms), whether to the extender (when directly connected to the extender) or to the router (when directly connected to the router).

However, I am starting to think that either we have several issues here (WiFi chipset vendors with this specific issue include Ralink (maybe, I cannot bring my desktop closer to either the extender or router), Intel, and Qualcomm (have yet to check with Realtek). And both the router and the extender utilize Mediatek chipsets (Xiaomi Mi 3 Router, running Padavan, and TP-Link RE200, running stock), or we have a completely new strain here (I am currently trying to test whether the network slowdown are related to system instabilities that are associated with Intel and NVidia graphics (testing out preGCN AMD graphics as well as Intel graphics, as that is what I have currently)).

I can add now, though I was planning on focusing on this later, that BT latency has deteriorated on the Intel 9260, with my sister's BT speaker having noticeable crackling and skipped audio with it ever since these issues have begun.

nah i'm not a fan of wifi extender, it´s better to connect directly because of the possible delay.

I'd rather not use the extender either, however we have a large house that was built around 20 years ago, and the walls are concrete, so ever since I had moved rooms, I can barely connect to the extender (let alone the router, I'm on the opposite corner of the house) without physically moving my PC across the house.
Hence why my current tests are either in my room, with either the desktop or laptop, or elsewhere in the house, only on my laptop.

Yesterday, I had tried to force the newer firmware version be renaming the files, however iwlwifi is refusing to work with the newer microcode, since the driver is coded to only use version 38, so I had instead renamed only version 38 (with .bak), and now it runs on version 34.

I can say now, that the dropping issue appears to be partially resolved, though I need more testing to verify this.
As it is, it looks that the speed is still dropping while under use, but it is no longer just dropping. but rather it is jumping all over the place, while half the time it remains at the speed it should be at, with our Internet speed not being affected, for the most part (due to our ISP's old infrastructure in our area preventing us from getting proper speeds with which I can properly test this).

As for the freezes, they seem to have disappeared without relation to iwlwifi, since 4.19.2, however with them being extremely unpredictable, I cannot verify this at all.

Thirdly, I have tracked down a small part of the freezes not having to do with hardware at all, but rather with a memory leak in Firefox, when YouTube videos are running, even if paused, regardless of whether the website is open or if they are embedded (it appears that Google is warring against users who are not using their new YouTube design, which itself, while for the most part I like it, has been causing performance and usability issues across the board for me, even on Chromium and Vivaldi (though not as bad as on Firefox and Gnome-Web), so until that is patched up, I won't be using YouTube for testing, but I'll look for another source which I can use for testing (something with 720p-60Hz preferably).
This leak is not Solus-exclusive, as it happens on: 1. Solus-packaged Firefox on my sister's laptop, 2. Firefox nightly on both my laptop and my desktop. 3. Firefox-stable on Windows 10 on every PC in my house, desktop and laptop alike.

In short, I think that the iwlwifi issues are not because of the firmware at all, but rather have something to do with the kernel driver, as I have been having similar issues on the QCA9377 (though I cannot verify this, since it is an unsupported hardware revision, hw1.1, as opposed to hw1.0).

Lastly, I have been having a new issue, on both WiFi (regardless of the card involved), and ethernet (RTL8169, using the RTL8111/8168/8411 driver), where it takes NetworkManager a long time to recognize network changes, such as enabling/disabling hardware interfaces, connection/disconnections, and I have to completely disable and re-enable networking if I do not want to wait (say, 1.5 minutes for each change).
Could this have something to do with this?

I had just finished my testing (I had to redo it several times due to kernel updates).
Laptop: Lenovo IdeaPad 300-15ISK, Kernel 4.19.7.
Router: Xiaomi Mi 3, running Padavan, up to 300Mbts@2.4Ghz, up to 866Mbts@5Ghz
The results:

iwlwifi. Local speeds drop immediately when in use by 30Mbts, and after every ~1-2 additional minutes of continuous use ,they continue dropping by 30-60Mbts, until they reach 40% of the max speeds (I had tested both 2.4Ghz and 5Ghz), upon which they either continue dropping (this time to either 54mbts, 24Mbts, or 12Mbts) or return to %60-%100 of the max speeds, and the network speeds becomes unstable (as in, the it keeps alternating from anywhere between 12mbts and the max speed).
Internet speeds take a hit (anywhere from 20%-80%, there is no fixed rate) when local speeds lose stability.
Additionally, reception randomly drops and rises about ~3%-24%.
When the network usage returns to idle, it takes ~15-30 seconds before the network speed stabilizes.
Network cards:

Intel Wireless-AC 9260, up to 300Mbts@2.4Ghz, up to 1.73Gbts@5Ghz, BT5.0.
Intel Wireless-AC 7260, up to 300Mbts@2.4Ghz, up to 866Mbts@5Ghz, BT4.2.
Intel Wireless-AC 3165, up to 150Mbts@2.4Ghz, up to 433Mbts@5Ghz, BT4.2.

ath10k. Unable to provide proper metrics due to my QCA9377 being revision hw1.1, when the newest supported by linux-firmware is hw1.0.
Reception goes down by the same as with the Intel 9260 on iwlwifi, but happens less often (and is overall slightly more stable), with the same impact on internet speeds when it happens).
Network card:

Qualcomm Atheros QCA9377/Dell DW1810, up to 150Mbts@2.4Ghz, up to 433Mbts@5Ghz, BT 4.1.

rtlwifi. No stability issues whatsoever, except for random ~1-3 seconds network hangs, network speed is abysmally slow in comparison to the others, but no different than it had always been.
Network card:

Realtek 8821AENF, up to 150Mbts@2.4Ghz, up to 200Mbts@5Ghz (claimed to be up to 433Mbts, however it had never gone above 200 for me, even on Windows), BT 4.0.

Interestingly, all BT audio issues have disappeared since kernel 4.19.7.

No graphical issues either, with any of the cards.

The situation seems to have turned for the better, though I am still facing instability on many days (some days I get ~1.6MB/s out of the ~3MB/s expected from the extender, sometimes ~1MB/s, but most of the time ~400KB/s, with semi-frequent hangs when downloading large or multiple files (i.e., torrents or system updates).
From speaking with others over at the the Solus-Chat IRC channel, it would appear that this may be due to issues with power management, even though disabling TLP adjustments does not change anything for me.

Update: The situation has been reversed, instabilities right and left.

The situation has barely changed as of yet, however the instabilities appear slightly later on 5Ghz (after ~30 seconds, as opposed to immediately on 2.4Ghz).

Additionally, it would appear that we have an additional issue with the internet connectivity, which is unrelated to the instabilities, hence the ~400Kb/s.