Thank you Andrew - some great questions.
With my introduction to Professor Lakoff (https://georgelakoff.com/), being a science guy this field fascinated me - how a person may react to a message based on the terms used, context of the message, before even processing the actual message conveyed by the message. Professor Lakoff describes this problem, particularly for Progressives. One can read his books, study how to craft one’s on messages, and remember these principles with every post, article, message. It is hard to be an expert for every topic and many people simply do not have time to study these principles.
On the flipside, we have Dr. Luntz, a noted pollster who famously (infamously) crafted a lexicon for Conservatives. His argument is that it is not what you say, but what people hear. Word/phrase choice is key. He conducts realtime research using those response gauges (audience provides feedback while receiving some message, such as a speech or a pitch) and sees the effect of different words. This has led him to produce a set of words that shift/bias/skew thinking around a topic irrespective of the actual message. He provided coaching and word choice guidance to Conservatives in the “Contract with America” - and the general discussion now has mostly adopted this language which, if Professor Lakoff is correct, puts progressives at a disadvantage right out of the gate.
Energy exploration instead of drilling for oil climate change instead of global warming are example “substitutions” that some attribute to Dr. Luntz.
Earlier this year, Forbes published an article about social media trends in 2017 - one of which was for a company to encourage its employees to discuss the company, its services, its products, in the employees social media. The company could publish all sorts of guides, recommendations, word lists, and hope that the employees follow them, should they even have the latest and greatest list.
There are often people who like a company/product and write about it in social media - if the company can help these people in using a messaging strategy, then the company is able to leverage widespread and consistent messaging instead of the current diffuse ad hoc messaging that may not synergistic or focused or helpful to the enterprise.
A lack of collaborative shared messaging that is able to be easily incorporated into all forms of communications is, at some level, a problem for everyone who cares about messaging.
I could imagine a news organization having internal style guides to help remove certain biases - a version of this technology may help even in more “neutral” communications.
On top of this, I know because I hear (maybe I am “tuned” by my confirmation bias to hear) many people on the podcasts I listen to contribute a piece of the puzzle here and there - often a certain insight about a “better” word/phrase choice. If only people would learn/remember to use these insights…
Regarding a concerns about the technology, it is a concern on many levels. I see this project as a basic technology that may be used for good or ill (like a telephone or Internet) - who knows if it will be adopted and useful, and if so, whether it will be used for nefarious purposes. So yes, it is a concern and it could be used to skew messaging.
Some say that has already happened by conservative discourse - a properly implemented messaging solution may help to counteract this and let people hear the message.