Obviously if your report makes assumptions about your data which aren't true then you might get invalid data out. I absolutely agree with having a single point through which writes to the data store must pass which enforces business-level constraints on the data. I just don't find SQL a convenient form to express those constraints (and my experience has been that any given business domain will have some constraints that are too complex to express in the SQL model, forcing you to resort to e.g. triggers - at which point the constraint is not integrated with the rest of your data model, it's just a piece of code that runs to check your inserts, which you could do equally well in any language); I'd rather do it in a "normal" programming language.

I see what you're getting at with reporting now, you're talking about doing actual calculations on the data? For mongodb I'd probably use its map-reduce API, at which point you're writing javascript and you can do anything, and performance should be fine. Though honestly other than performance I've never had a problem with just fetching the rows I need and doing whatever transformation/aggregation logic in e.g. python. SQL has never struck me as particularly typesafe or gaining much advantage from being close to the data; its sets-oriented way of thinking can be helpful in some places, but it's not the only language that can do that.

If you like SQL as a language for writing logic in I can see why a traditional database would appeal. But even then I feel like input validation, data storage and query processing should be separate layers (and I see some movement towards that with e.g. MySQL offering an interface to do key/value lookups directly). If SQL is a good way of doing transform/aggregation logic then it should stand alone, and I'd expect to see SQL-like query layers on top of key-value stores.