Gays for Trump never had much prominence and you speak as if they make up a sizable chunk of a convention of 20k bigoted people. 3.8% percent of adults in 2016 identified as lgbtq and an estimated 14% of those voted for Trump that year. That means statistically there would be 99 lgbtq people there out of 20k, but:

On the night of July 17—the Sunday that check-in for the convention began—there was a 120 percent increase in users online in the area, more than all previous evenings in July. As of last Wednesday, a representative from Grindr tells Broadly, “more than a thousand guys have used the app near the Quicken Loans Arena.” The total attendance of the RNC is estimated to have been roughly 20,000 people, so this data suggests about five percent of attendees used Grindr during the convention.

If their data is to believed, that means that the stadium had over 10x the predicted amount of conservative lgbtq folks relative to the general population and survey results of preferred parties by noncloseted individuals and, in fact, had not only more gay individuals than the general population by percentage (which seems crazy for 14% support), but that they all used grindr, an app with less than 3 million global daily users at the time. It also doesn't filter out the few women at that sausage fest.

It would appear that grindr has decided in this environment that it's better to keep their head down. Obviously a typical major service like grindr uses cloudflare. They are not going to "crash" as a result of overusage in an area because any concentrated traffic is easily spread across their network and overloading traffic is handled by cloudflare seamlessly. Unless grindr is actively giving location Hotspot maps for every event I think any dismissal of the article's evidence is in service of narrative rather than evidence, ironically.