A recent article in Phys.org talks up the ever increasing stability of perovskites, yet still their main problem is that they deteriorate in the sun. And doing well in the sun is the most important thing for a solar panel.
The promise is that perovskites which have already reached 22% efficiency can cover different band gaps. i.e. conventional solar cells use violet-to-red wavelengths of light and they're really stable and getting cheaper and cheaper. Where perovskites could come in is as another layer on top of a silicon solar cell or a series of layers. The problem is you wouldn't want to put something unstable on top of your silicon solar panel; one of the benefits of solar panel is they have a potential 40 to 50 year lifespan. What if adding a layer of perovskites to harness another 5% or 10% of energy then caused that panel to completely or almost completely fail because the top layer breaks down and starts to reflect light away or absorb it in such a way that it doesn't cause electrons to bump along.
Perovskites are promising but at this stage HIT technology developed by Sanyo (and later taken on by Panasonic who acquired them) has now had its patents expire, so large solar majors are looking to use the HIT technology to add a few more percent efficiency to existing solar panels. Commodity Mono PERC panels which are already on the way to 360W by the end of 2019 could then easily become 380 or 400W; making commodity panels competitive with the world's best panels namely Sunpower.