Chris.A.G wrote 

Base ap previously fed 3 APs, with -72 to -77 db connection strenths the weakest of which resulted in TX of around 117mbps , RX of between 117-234mbps

-72dBm to -77dBm is not a »strong« signal, it's a pretty weak signal  with packet loss occuring at around -77dBm.

For a stable mesh link you need much higher values (note that »higher« means lower numbers for negative values). The signal strength between two mesh nodes should be better than -60dBm (for example, -55dBm or -52dBm). Note that dBm is a decadic logarithmic unit for the attenuation of the signal, that means if your signal strength is -72dBm, it is 100 times weaker than a -52dBm signal.

A free line of sight (not »line of site«) is always required for a good signal between two wireless devices, it is not a special feature. Any obstacle in the line of sight between two devices will cause interference which will significantly decrease the signal strength. Even air, rain and snow causes attenuation, but trees and hedges etc. must be avoided at all costs.

What are the distances between the mesh nodes and the root node (in meters please)?

Chris.A.G wrote

Yesterday was a cloudless, clear sky, and the leaves are falling so I would have thought that it would be getting a better connection.

 

Usually signals are be better in winter than in summer, but it depends on so many environmental parameters that it is difficult to establish general rules. For example, in my country we currently have high humidity – today over 95% – and this can attenuate signals, too.

I know im pushing the limits of distances, but its a rural property with limitations on where I can put in APs (short of trenching and running power lines, but there are certainly costs and challenges going that route... at that point, would need to decide if simply getting another internet connection for the  problem branch would make more sense). 

According to the TP-Link FAQ for mesh networks 180m is the maximum recommended distance outdoors with a free LoS, but it's also recommended to not exceed -62dBm (see Q6-3 in the FAQ). Note that the 180m is for U.S. models, for EU models I would not even try to cover distances more than 100m (we have much lower limits allowed for the signal strength).

What you can try:

• If possible, connect the cams to the 2.4 GHz radio and use the 5 GHz radio exclusively for the mesh backhaul.
• If you use controller v3.2, set WiFi mode to 802.11ac-only.
• Try different fixed channel widths for 5 GHz: 80 MHz, 40 MHz, 20 MHz.
• Always check signal strength after changing settings.

If this does not improve signal strength, try to reduce the distance between mesh nodes, i.e. use longer cable for the PoE supply ad move the antenna pole nearer to the root AP. Except for the above measures there isn't much you can do if the signal strength is weaker than ~-62dBm.

I all this doesn't help, consider use of directional wireless links, e.g. with Pharos CPE510. I always use CPEs for long distances > 100m. But CPE510 is limited to N300 WiFi rate, that's max. 100 Mbps Ethernet throughput. Alternatively, CPE710 is a new AC867 device matching WiFi rates of an Omada mesh. But I haven't tested CPE710 since it's still not available here in my country.