The UK as a whole is nominally using a threshold of 20 infections per 100,000 population in the last 7 days as a benchmark to decide whether to remove a country from the exempted list. They also take into account other factors such as whether case numbers are rising or falling, number of deaths, testing regime etc. As far as I know Greece has not exceeded that figure so Greece has not been removed from the exempted list. If you look at the 14 day figures published by ECDC
the UK is at 24.1 while Greece is at 26.5 so not a big difference. Portugal are at 34.5 on 14 day figures and they haven't been removed yet although it is under consideration. That gives you a rough idea of where Greece stands.
Scotland and Wales are also taking decisions based on specific incidents, in this case local outbreaks which traced back to specific flights from Greece. In the case of Scotland they have applied the rule to the whole of Greece while in Wales to a specific island. Whether UK tourists planning to come to Greece in September and October will appreciate that nuance is debatable so how many the Scottish and Welsh decisions will deter from coming is unknown.
The sharp rapid rise that started in Greece in mid July and continued until mid August seems to have flattened off quite a bit in the last two weeks. Whether that flattening will continue or turn into a fall is anybody's guess. Most Greeks will have gone home from holiday so that should tend to reduce the spread but schools are going back on 14 September and my gut feeling is that it will cause a rise. How big I have no idea. An expert from Sage was asked about UK schools going back on Sky News two days ago. Her reply was that the virus spreads wherever people gather and mingle together and schools fit that description. She also said the key to limiting the spread was test-trace-isolate and the UK system was not yet good enough. I am pretty sure the same can be said about the Greek system.
PS People like Matt Hancock bandy about tons of figures to prove that test-trace-isolate is a success. The purpose of test-trace-isolate is to reduce the spread. If the spread is increasing then by definition test-trace-isolate is not working. The percentage of people traced is irrelevant if it doesn't reduce the spread.