* there's no such thing as 'unstructured data' (if it's unstructured then it might as well be white noise)
Facile. There's a world of difference between ambiguously-structured data, such as natural-language text or images, and unambiguously-structured data like the values in an RDBMS with well-defined columns. Quibbling about the ideal meaning of "unstructured" is unproductive.
* it's daft to call a whole family of databases"NoSQL" then spend the next 3 years building SQL on top
Inasmuch as "NoSQL" was backformed into "Not only SQL" shortly after its introduction, this objection is also baseless.
* taking data out of a relational database and putting somewhere else doesn't make the data "non-relational"
Data itself is never relational or "non-relational", so this is vacant. Data can be in a relational structure, or not; if it's not available in a relational structure1, then it's not available in a relational structure.
* just because the security cameras in you HQ record more data every day than your customer ordering system, it doesn't mean they hold more value
Straw man. No one cited in the article (even Microsoft's "envisioner") claimed otherwise.
* when it's harder to merge data in your database than it is using vlookup in excel then you probably don't have the right database for analytics
Perhaps, though that's such a vague claim it could mean almost anything.
I'm all for skepticism and generally even for contrarianism - the IT industry, broadly speaking, is far too fond of its fads. But this sort of "I work in a narrow part of this field and it's all a homogenous lump where everyone's problems are the same and no solutions other than mine are valid" crap is both tiresome and useless.
1And necessary normal form, modulo denormalization with controls to ensure a valid transformation to normal form always remains possible.