Just to be clear, *.epw and *.ddy are simply file formats used by EnergyPlus. You could conceivably put any weather data into those formats and EnergyPlus would run (barring reality checks in the program for values outside of physical possibility...). Whether you will get meaningful results, though, is a different question.

If you just have typical day information (hourly values or averages?), how would you expand that to the other days of the month? You can just repeat them, but the weather file in my opinion would be too unrealistic to provide meaningful results. You also did not mention solar radiation, which is a main driver for cooling loads.

No, my recommendation is that you don't try to turn such sparse data into an *.epw, but that you go find a real weather file for your location of interest and use that. That would also be a lot less work, since creating an *.epw is not something that can be explained on a bulletin board.

The *.ddy files supplied with the *.epw files on the EnergyPlus Weather web site are taken from the ASHRAE Handbook of Fundamentals for the same WMO station (if available). If you know the lat and lon for your location, you can find the closest HOF station in the database (see PreProcess/WeatherConverter/ASHRAE2013YearlyDesignConditions.xls) and use that. Having said that, I feel there's an over-reliance on the ASHRAE design tables. There ARE different design procedures in different countries, and real engineers actually tweak the values to meet their needs.

Yes, you do need multi-year records to calculate design conditions, but these are not so hard to obtain anymore. Perhaps a bigger barrier now is that people lack the software (or don't have enough interest) to extract the design conditions, which really isn't that hard.