Mann,
The article actually has a quid each way. The water layer may be too thick or it may be too thin, yet the probe capability was built on known physics and chemistry and there is no indication that the laws of physics are different on Mars. So the more likely explanation is that the H2O molecules do not penetrate the surface.
Facts:
We know that the probe went into regolith not permafrost/ice. That layer is hard, resisting intense scraping and rasping
We know that the permafrost remains intact with only a few centimeters of regolith protection implying extraordinary thermal inertia properties.
We know that the probe can detect minute traces of free H2O molecules and that it can find none in the regolith.
We know that the regolith was disrupted and the probe open to the atmosphere during the only measurement where water molecules were detected.
I am sure that they will have spread the probe times to encompass the daily spread. And Mario's experiment although an innovative approach that he should take a bow for is as an analogy, like putting a 10 gram weight on a shipping container weighbridge and claiming that since it did not register it must be weightless.
The absolute dryness of the regolith is an unexpected finding and an incredibly important outcome. (dryness in this context being a relative term as their expectations of molecule thick films would have been classified as waterless in any other context). But this lack of any H2O molecules obviously makes some people uncomfortable and confirmation bias rears its head yet again.