Opening the data (along with the tools made for it) up to other specialists is definitely the right thing to do.
And it worked: They had a nice new observable they defined and rather than having to go to LHC people to check the data, they could do it themselves.
Things like a prof from one field walking past a blackboard with some formula or other for a different field and noticing that it is remarkably similar to something he knows intimately, or a solved problem in his field.
Similarly fresh eyes may notice or question something that trained eyes has overlooked virtually by habit.
That said, there is a risk that one get a "eternal september" scenario where any useful input is drowned in the flood of noise from a stampeding horde of newcomers.
You limit access to the data, you limit ability of others to make an important discovery. Status quo won't offer free access to this data because it apparently reduces their importance.