YES! This is it! Why it happens?
My only suggestion as of now is a dust. I just can't see any other explanation to it! Try to print with it with significantly increased cure times. It should print, but model still will be kind of bad.
Specially to test output of projector we need a spectrometer (spectrograph). Of course it costs some 2-5-10-100 grand$

But I found a simple solution to make a spectrometer at home. Here is instructions.
Buy simple: http://store.publiclab.org/products/desktop-spectrometry-kit
DIY: http://publiclab.org/wiki/foldable-spec
http://publiclab.org/wiki/dsk
http://publiclab.org/wiki/ottawa-homemade-spectrophotometer
I encourage everyone who is interested to do the same so we could compare data and analyze. Likely that projector started to be somehow dimmer in UV-blue region of spectra.
1.6 seconds. That's unbelievable. Even MJR G+ is not giving me such a cure times...