I don't believe this is true. Applied Physics Express just published a paper last week from Corning entitled "Impact of Carrier Transport on Aquamarine-Green Laser Performance". It details their work on evaluating green and almost-green laser diodes, and the work was done quite recently. And based on a job opening or 2 that I know about at Corning, I don't believe they're stopping this project.
It's possible, but not necessarily directly causal, if you know what I mean. Nonpolar doesn't automatically mean better beam specs, but it allows you to do other things in design to get better specs. It's thought that gain may be higher on other planes, so threshold is lower, which is always good, more power at the same current/same power at lower current. The lack of polarization means you can go to wider quantum wells, which can increase your confinement factor and lower the threshold current as well. The increased confinement factor means you can possibly change your transverse waveguide structure (cladding layers, SCH, etc) as well. So there are several ways you can possibly get more power out of a laser.
All these improvements lead to latitude in your design, and the possibility of being able to use things like narrower ridges. If you can get the same power out of a narrower ridge, all else being equal, the divergence can be better.
Make sense?