You're telling me that it is simpler and clearer to write something like
$Sentence.Myself.Subject()
$Sentence.Verb("to be").Present.Person(1)
$Sentence.Negation.Absolute()
$Sentence.Verb("to convince").Participle.Past()
$Sentence.FinalPoint()
ConsoleWrite(SentenceBuild($Sentence) & @LF)
rather than
ConsoleWrite("I am not convinced." & @LF)
That may fit your mood or models but I dare to insist that this is counter-productive.
A non-trivial application can be a typical simple invoicing system, for instance. There you're likely to find a table of clients, a table of delivery locations, a table of invoicing addresses, a table of orders/invoices, a table of items, a table of items ordered (some delivered and some still pending delivery), a table of paiements associated to orders and invoices, and most probably some more tables to fit the real world (suppliers, stock management, ...)
I don't buy the idea that such a common application can use 90% of flat queries.