You're right, there is value from SQL over event data, however, I feel like it's a missed opportunity to simply apply the same paradigms to a different type of data. I'm not suggesting that SQL be thrown out but a new language needs to be available specifically for event data.