Dear community !

However im long therm MT user, ill post this topic under this category, since i feel like a total beginner now. I created this topic just to understand "what is going on" in my home and pour ot my heart... I appreciate any idea, workaround, explanation, response if you have time and mood. I will also accept that you will consider me a fool, but at least I will know that the problem is "with me".
Couple of years ago, i replaced my old RB951 with hapAC3. I was pretty impressed that time with the wired-wireless speed of my new device. As time has passed , technology/competition has advanced and i encountered some issues with new Apple and IoT devices i decided to swap to RB4011+RM and also benefit from "160MHz Wireless network". This was a huge mistake and after releasing the new Wifiwave2 drivers even bigger mistake (because of known lack of 2.4GHz support here). However i always prefer cable connection, demand on stable and fast WiFi in our household is always larger. I sold my RB4011 and started to search for the right replacement. As i wrote on different topic, i have a spare unused hap .ac2 (in case of something...), so ill started with fresh ROS (7.10/now 7.13.5) with Wave2 drivers. I was totally impressed by the coverage and speed of my wireless network !! I have great coverage in comparison with my previous mentioned devices, with insight, steadily and always achieving speeds between 520-560Mbps on .ac wifi with this device, approx 6-7metres away from the router, what is in our 75m2 apartment maybe the furthest place to "use Wireless connection", except other rooms "behind the walls" ! After that, i immediately borrowed back the .ac3 for Wave2 test, used exact same software setup, placed in exact same location and made some readings from the same places. The 2.4Ghz network had the approx. identical performance, but 5ghz was significantly worse ! This means, in place where i reached 200-230Mbps on .ac2, i had measured around 60Mbps with .ac3 (because of lack of 5ghz coverage). It was a disaster on 5Ghz in every single place ! Time for ax3 and ax2. Both of devices performed in my environment absolutely identical, no mentionable difference noticed. I was able to get approx 600-620Mbps, but only 1meter away from the router ! When i made measurements like i did previously, ill get around 100-120Mbps lower results with most recent MT ax devices as on .ac2, same distance, measuring point, setup, client device !
I dont understand what is going on, and i am convinced that I have not made a mistake. Every single device i tested was installed with NetInstall as a "fresh device", with Wave2 and absolutely same configuration. "Im pretty happy" now with .ac2 performance, since this is something what I haven't experienced till now with MT Wireless, but we all know that this device isnt future proof... Actually it isnt "tomorrow proof". Running with 650kb free space, rejecting any future updates (to avoid issues during updates because of lack of space) isnt a good feeling, also the "WiFI5" standard is maxed out yet.
During "research" were present also different .ax devices like Synology RT6600AX, ASUS EBR63 and Linksys MR2000 to have some comparison with other brands and "everything was done in couple of days". Used HE Network Tools (iperf3), WiFi SpeedSpots on iOS, Speedtest...

Friends, do you have any rational explanation for the described mystery and facts, as I have tears in my eyes already... ?

P.S. - Thank you for reading this far
P.S2 - Please, write somebody ROS image into RT6600AX and take my money !!

I started with CAPACs, and then I migrated, to tplink business access points but kept MT for excellent routers.. AX3 is an excellent router ( and you should get decent local wifi near the unit ) and an even better router is the RB5009

Did you note the frequency/channels used on each device (ac2,ac3,ax2,ax3) when performing your speedtests?

And my experience: if your wifi signal has to penetrate a brick wall (even if it is only like 10-12cm commonly used as interior wall) - forget about MCS11 at least. So no benefits of 802.11ax in regards of bandwidth. But you can still benefit from all the other 802.11ax specs. So it depends.

I dunno for sure with AX whether it's a good idea... but I've always used 20Mhz or 40Mhz if I cared about distance from Wi-Fi. You'll lose max speed near router, but I suspect speeds at a distance would improve/more stable. Playing with channel selection certainly be a good idea... although at 80Mhz you don't have many choices .

It looks like the ac3 is broken because it has a 5.5 dBi antenna compared to 2.5...

Thank you very much guys for responses ! I really appreciate it !
As i wrote, the configurations for ac2,ac3,ax2,ax3 was really the same. Honestly, there is really not too much to elaborate on Wave2 drivers in comparison with legacy drivers. Fixed channels, same channel width, same security profile, same DTIM interval (2412,20/40Ce ; 5180,20/40/80 ; WPA2/WPA3). So absolutely no difference in setup on all devices. The only change was of course the standard ac vs ax (2Ghz N/5Ghz AC vs 2Ghz AX/5Ghz AX), but ive done tests also with max ac standard on ax devices, just to compare them with mentioned .ac2 performance. The result was the same. Router is located in living room, clear direct view to kitchen without obstacles/walls, straight line, approx 6m distance : ac2=520Mbit, everything else (ac3,ax2,ax3)= 400-420Mbit. The ac3 performed even worse. If i get closer to the router, living room sofa, clear view, straight line, approx 2-2.5m distance : ac2=560-590Mbps !! The ac3 performed here also worse, only the ax devices could catch "near speeds". Ive got higher speeds on ac devices only if i get even closer, so 1m to the router what is useless... It means, no device in my household can perform at least equally to ac2 if i move further as 2metres. The speed drop because of lower 5Ghz rate/switch to 2.4ghz is in other rooms (for instance bathroom) even worse, as i wrote 220-250Mbps to 60-100Mbps ! The second room (behind the wall) is approx the same with ac2 and both ax devices, the ac3 is again "nothing here". I will agree with at least same performace/speed as my ac2 delivers, but i could be at least "safe" because of HDD space and have a future proof device. When i tested RX6600 in the kitchen, so the "6metres" distance test, ill got 740-760Mbps speed This was the lowest value... Yes, i know, 3x more expensive device, 160Mhz etc..., i was just curios if there is issue "around me", or it is really something strange happening with my MT devices. I would immediately swap to Synology, but their "SRM firmware" is so much "user friendy", that you cannot setup some basic things like two SSIDs for same subnet, and the other options are even worse, but this is a different story...

Did you check the transmit power of the radios while you were testing those APs?

@neki - thank you very much for your response. Honestly, no. I was checking the signal rate of each fixed client and the speed. Reason was what i wrote, fresh ROS via netinstall, same ROS version, everything fixed. 4different devices from which 3 performed worst as a low cost entry level internal antenna device couldnt be "buggy" in case of software transfer limits with same country, frequency set (manually). At the other hand, you are right. The focus was on same setup and real comparison, i didnt checked the background because of upper described facts.