I am not sure why the nanoplastic levels are so much higher than the microplastic levels in these plastic cups.
I'm picturing a plastic surface as being made up of many polymer chains linked and tangled, but also with many smaller chains not firmly attached. There would be some clumps of bound chains, which would be easily freed to become microparticles, while the individual chains would become nanoparticles. Heating a surface would flex some of those chains, releasing new particles. Environmental factors would constantly break off or expose new loosely-bound particles. Whether that picture matches reality is a different question.
It could be that researchers are using different measurement methods for these particles, which then results in wildly different particle number figures.
That's my guess too. With any new area of investigation, there's the problem of lack of standardization, both in methods and terminology. Two studies reporting nanoparticle counts might be comparing apples and jet turbines (ie. not even remotely close). That's why I'm thinking that the amounts found in that study will not be shockingly different from many other everyday exposures, once some standardization of measurements arrives.
What seems clear is that hot food or hot beverages in contact with plastic results in very high levels of plastic particle release, compared to other sources.
Not necessarily clear, since "those other sources" may not have been measured in a comparable way. Maybe wiping your mouth with a synthetic fibre cloth or drinking cold water from a new plastic cup releases a comparable number of particles as the microwaved items in the study. We'll have to wait for a proper study of everyday exposures.