you highlight a good point. It is in fact often the case that data are published in different formats (and sometimes also licenses) depending on the temporal or spatial granularity. In the case of air pollution data, we discussed both cases (hourly as well as daily pollution levels).
To answer your question: you should base your answer on daily average values of air pollution, i.e. air pollution levels for the past 24 hours. This has following reason:
We based our definition of important air pollutants on air quality standards (primarily established by the World Health Organisation). These standards prescribe different update frequencies for pollutants. For instance O3, PM10 or others have to be reported within 8 hours or 24 hours (as I can see, in Czech Republic PM10 is reported every 24 hours). These update frequencies are not fully agreed upon among countries. Some countries choose to measure the same air pollutants in different time intervals (see for instance here, and here, for a comparison between the EU and the USA). What is published as open air pollution data therefore depends on how air quality is defined within a country. Therefore we decided that all air pollutants should be published at least once a day. Publication of daily means is our baseline to compare across countries, and the main criterion for us to score a submission.
But please do use question B3 and B8 to document that you also found hourly air pollution data, as well as the formats in which they were provided. This gives valuable extra information and helps everyone interested to understand the data in Czech Republic, and what the data can be used for.
I hope this answer is helpful.