I think you are being a bit optimistic there, with many of these designs needing two power conversion stages (with the need for power factor correction etc). Tesla claim only 91.5% for the V2 supercharger; I haven't seen a figure for V3, but V3 is definitely a two-stage design.
Also, your measurement probably has more losses outside the charger itself. At the very least, you have the cable/connector losses which won't be trivial with the chargers located in cabinets some metres away from the charging stalls. Some high-power CCS equipment has active cooling of the connector and cable (so losses for the cooling gear, but also pointing to very significant losses in the parts that need cooling), though I am not sure if either Tesla or Ecotricity are using that.
If the Tesla-end measurement is at the battery terminals, then you have all the vehicle auxiliaries (coolant circulation, control electronics etc.) to add. If the Ecotricity measurement is literally the electricity meter reading, then it includes cooling auxiliaries for the charger (fans, probably some measure of liquid cooling?, control gear).
So while I'd agree with you that 87.3% has room for improvement, it's not particularly surprising, especially if some of these external losses are included.
