Having used Qlik Sense on a daily basis, the shortcomings with UI/UX can be overcome by creating your own D3 visualizations and custom mashups in JavaScript. They also have an open source community which is slowly closing this gap.
The power of the product over its competitors is really the in-memory API engine. I think they should now just focus on getting Qlik Sense enterprise ready as the server deployment still needs improvement.
It sounds like whoever built your QlikView dashboard may have not known what they were doing. Maybe they just loaded in the data without doing any data modelling. QlikView is extremely fast because all the data is held in-memory. It will always beat querying a database on disk or Tableau which holds data in a mixture of on disk and in-memory.
The query language you mention is used in the UI, but you can use full SQL queries in the loading of data into the dashboard and also for creating a data model.
Yes the visualizations of QlikView are now out of date, which is why they have released their new product called Qlik Sense which looks good, modern and completely web based.