I would love to answer this, but I have no idea what "strong linked data" means. These techniques have been used in a lot of products, the dbpedia.org system has been used by a number of systems, ranging from Watson (maybe IBM is considered too academic) to Siri (I don't think Apple is an academic group). Schema.org and Facebook's Open Graph Protocol are also big users of linked data vocabularies and web linking schemes.
Going to government, there's been a lot of work using linked data in various ways. The Brits are the lead, a number of their open sites, based on the Ordnance Survey maps among others, use linked data. Within the US, we have demonstrated a lot of uses at hackathons, in some of the competition winning apps, and in some the info sources available on line.
There's an article Jeanne Holm, George Thomas, Chris Musialek and I wrote that covers some of this - IEEE Intelligent Systems and some thoughts about what we are doing at data.gov
So my fast summary is that like many technologies this is being used as a component in many apps, it is not by iteself proposed or fielded as a be all and end all - but without the URIs, the data doesn't make it to the Web, and then we cannot exploit many of the powerful things that open data allows.
(I have a column in the soon-to-be-released issue of Big Data called "Peta vs. Meta" that says a bit more about how the schema.org stuff is used - see also schema.org/Dataset)