If it is a change in water take up, it has to be because of a new chemical makeup in the inks we've used for years. It makes sense that it is an ink/water balance/chemistry potency, because according to the people I've spoken to, many companies are using these inks without any known problems.
I think it could be a new resin in formulation. Or maybe any other "eco-nail" for printing industry coffin.
From my experience many companies don't care much because of high dot gain and especially because of deformed dot on printed sheet. The answer for them is "compensation curve"
All print shops with whom me or my colleagues talked to do not see any problem in their high dot gain and they do not have microscope to see their deformed dot.
One reason is because same supplier do not see problem too. They intend only to sell ink and don't care how it work. Technical support is very poor. Actually most of the support is lots of talking to convince us that problem is not in the ink or ink/fount pair.
Other reason is that printers are still able to sell their printed sheets to the customer – position is "why bother" consistency?
Just today we've finished another set of tests with manroland servicemen who cheked again all rollers/cylinders/pressures/bearers etc.
And we've test different inks – with one of the ink we've obtain acceptable result.
Test method was simple as we've tired printing charts: we've output a linear testfrom with 70% all over the plate (200 lpi, round dot, 0 degrees).
With our "bad" ink there were a lot of big horisontal lines over the sheet from head to tail – like form rollers gone made and trying to beat the plate to death. Dots were distorted with ragged edges. That's when we're undestand that fount cannot clean the spaces well enough (tested with lots of water levels)
With our "acceptable" ink there were no lines or distorted dots on the sheet at all. I've finally saw a sheet 70x100 full of clearly printed 70% dots right out of our press. And they all were round, not perfectly, but good enough for wet offset.