I agree with you that the NYT's coverage of India is heavily negative, though I think you could reasonably go far less specific and argue that broad focused American media in general tends towards negative coverage
period (part of my major problem with it, "If it bleeds it leads" is not a new saying), particularly about international matters (because international matters are often shortchanged anyway, foreign desks have been reduced/eliminated, etc.). With that said, be careful of absolutes. I just went to the NYT's page for the topic of India, and yeah, it's massively negative, but just glancing through the last few months I saw:
1. "A Photographer’s Homage to an Architect’s Modernist City": http://www.nytimes.com/slideshow/2016/11/03/t-magazine/art/m...
2. "Mark Morris’s Love Letter to India, Through Music and Dance": https://www.nytimes.com/2016/10/23/arts/dance/mark-morris-so...
3. "How Do You Hail a Tractor in India? All It Takes Is a Few Taps on Your Phone": https://www.nytimes.com/2016/10/18/world/what-in-the-world/t...
That's just a quick look at recent news, and given that the NYT has to be feeling pretty down about the overall state of politics worldwide right now I doubt that's even a fair slice.
Honestly some of those are actually pretty nice/interesting pieces, which does bring up a point: I've long given up on making media more positive generally, negativity sadly seems to sell pretty well, so you have to go to alternates/aggregators that are better (including HN). But I really do wish there was at least a filter for "positive stories", a simple binary the NYT or WSJ or WP or whomever would use to broadly categorize stuff that was about /disaster/war/death/crime/terrorism/fights/eldritch horrors from beyond space time/ vs anything else. If you just go making a list of other stuff from a lot of media you might be surprised about the kind of interesting pieces that tend to get lost in the noise, and the important perspective on progress in the world it brings. Broad categorization seems perfectly doable there, doesn't need to be perfect to be highly useful in raising s/n, and I'd love if outlets would just do it themselves and see how their readers react to having a choice. I guess at some point ML/AI assistants might be able to do a helpful job based on word stats, but it'd be best if whomever wrote the article just tagged it at the start.