The "platform" is not the most laborious nor costly part of a financial research terminal such as Bloomberg (or others such as FactSet, Capital IQ, or Refinitiv). It is the data, its scope, quality, assurance and delivery speed, that consume tremendous labor and ongoing delivery costs. The data includes hundreds/thousands of curated fundamental data points covering thousands of corporations globally spanning decades of history. All the history is easily merged through these platforms to be correlated with current, realtime, global market feeds that are no more difficult to consume than it is to drink water from a fire hose. (Be careful not to underestimate the service level expectations of current market data consumers; you do not have the sluggish luxury of taking minutes, or even seconds, to deliver all assured data to all customers [as close as possible to] NOW.)
"Openness" is already a highly valued and well-supported aspect of financial research platforms. Users of those platforms typically integrate third party data sets as well as their own data in order to build unique investment models. So in addition to supporting common user scripting mechanisms such as Excel/VBA, they serve up their data through well-documented API's scaled to satisfy vast needs at high speed to be consumed through whatever technical toolset a customer may choose. (Python? Okay. Are you sure it's not a job for c?)
Good luck competing with those guys. None of them is resting on their laurels, and they already stand atop many years of continuous investment in these highly sophisticated platforms and their ever-expanding data sets. The products are only sustained through 24x7 global research, aggregation and delivery services. Though those companies have been highly profitable, they're also highly competitive and not nearly as nearly as fat in their pricing as the article suggests.