I've tested via the modem itself, PC and Pixel using both 2 Degrees and Chorus test servers.  All consistently do 65-66 mbps

MaxineN:

 

I can't find the LTE CAT grading on the TP-Link jobby so.. @SaltyNZ.

 

However with the speeds you're seeing... wouldn't be surprised if this is a testing issue OR FWA is also weighted pretty heavily... like the other two do.

 

As they are marketed as 'broadband' services - with unlimited data - you won't be shocked to learn that the relatively small number of FWA connections uses way more data overall than the far larger number of mobile connections. So, yes, FWA plans may be prioritised vs. mobile for traffic management purposes.

@SaltyNZ --Even for rural towers - where the mere presence of an unlimited plan implies spare capacity?  

Some background:-
I have Starlink Residential Lite which is about to be downgraded to Residential Lite 100.  There is also a new Residential Lite 200 which puts the "100" plans at the end of the line for prioritisation.

As I don't do gaming or video conferencing, the higher latency of FWA is not a concern and it gets away from Starlink's occasional bursts of high packet loss.  I thought that if FWA gave me the lower end of the DL speed of the mobile minus a bit for the FWA weighting then I might get a 100mbps end result at a lower cost.  I hadn't realised the FWA weighting was so extreme.

I am only a few days into the no-risk cancellation period so will continue to trial it.