How can I determine what speed on my devices I am actually getting?

Try testing using OOkla SpeedTest. There is an app

in both he MacApp Store and the iOS App Store

or you can access the website and use the web version.

The advantage of this one is that you can select from

a variety of different servers to test from since not all servers

will be full speed all the time and some will be consistently

faster tan others.

Also, do you have the mesh with multiple units or just

using a single MX4000? If multiple, back of and just test

with the master as close to it as you can get and then

start backing off and then start to add satellites for

optimum speed through out your area.

One last thing, what is your environment like

(urban apartment complex, suburban development,

or country environments)? Reason for asking is that

interference can wreak havoc on data signals. While

initial link connection speed is a simple handshake,

transmitting complex data back and forth in a high

interference environment can have many corrupt

packets and many retransmit requests back to the

internet source.

Degradation is to be expected in a mesh system because even though it’s giving a strong signal to the mini, it doesn’t have similar towards the router. So it buffers and throttles depending on what’s being thrown at it and how fast. By all means optimize as much as possible but lower your expectations. You might have already reached it’s limits…

Cheers

If you have a strong signal without the mesh units, you don’t need them. The mesh units take a medium/weak signal at the border of the main unit and beam it to extend it to the devices in the extended section. That’s not the same as having a strong signal from the unit, it’s more like breathing life into the signal at its edges so more devices can drink into it. In other words, irrespective of what the marketing says, meshing is a compromise and comes with throughput costs. There’s no working around tit (other than to wire your house for Ethernet, which if possible, do strongly consider).

Summary: Try removing the mesh units and test with the original strong signal. If it’s an improvement, stick with it, if not, use the mesh units but with lower expectations of throughput.

The Apple Extreme is since 2016, a discontinued, obsolete product, therefore insecure and best avoided. Stick with mesh until the rooms are wired for Ethernet. You can then connect those LAN points to routers in bridge mode which’ll get you the real thing point-to-point.

Yes, of course. Hard-wired lan feeds pure 1Gbps from the main router to the bridged routers in rooms which in turn offer pure WiFi AC to their assigned areas (pure because their backends are solid). Mesh is a compromise to re-beam weak WiFi signals from border points to extended sections. Since they’re working with weak signals on the backend (unlike LAN), their offered good signals to client devices are sort of fake (which get revealed upon testing as you’ve found out). Meshes extend range but not throughput.

There is a "mesh myth" that seems to be perpetuated by the industry

of high WiFi speeds everywhere (at least they don't say it's not true).

A satellite unit can only retransmit at the speed it receives, regardless

of how powerful it boosts the signal. So, all @hcsitas has been telling

you is very true.

The best way to achieve full WiFi speed throughout a home or any facility

would be with a main ethernet router wired with ethernet to WiFi

"access points" located at strategic points.

The whole mesh concept is more a convenience of setup than

it is of truly "high speed throughout".

I think you have it backwards.

700mbps is probably your LAN WiFi speed.

65mbps is the Internet speed to that particular speed test server.