@flymetomoon 

The max you will get out of the LAN ports is 1000 as that is the max port speed.

The max WIFI speeds may be due to the device.

See https://www.duckware.com/tech/wifi-in-the-us.html for more info on WIFI speeds

I don't believe that the AX11000 speed test from the management webpage is accurate at higher speeds.  I wouldn't trust what it shows you.   See https://community.tp-link.com/us/home/forum/topic/224486 My internal speed test from the router says my WAN connection 600ish when my clients actually receive 900+ when I run speed test from a wired client. My subscribed service is 800. Because of that I set my QOS to standard and manually filled in the values with my subscribed speed.  Because of the inacurrate internal test, gig ports on the router and wifi client issues mentioned above, you're going to have a difficult time testing if you're really getting the 2.5 speed.

@ArcherC8 @djf5502

Thanks for replying. I have added the pictures for better illustration. I suppose the link aggregation on LAN port 2&3 can achieve 2Gb/s but I cannot see any improvement. It is weird that the AX11000 got 100Mb/s more than 1Gb/s.

@flymetomoon 

My two cents:

As far a router giving you speed readings, I prefer to use a standalone program/link to give me a better real world reading that includes my device, Ethernet cable, and the LAN port, even then sometimes I wonder about the speed. For example, your router may be connected at 1.5G using your super high speed WAN port, but your router has only 1G LAN ports so the max you can get will be 1G from your PC where your router has 1.5G. 

I have not used link aggregation but what I have read is that most types/setups improve throughput so multiple devices have multiple paths through the network. With a single device, I can see where it might even make it slower with the added overhead for a single device.