I go along with the stated 6 hydrogen atoms per cubic meter volume as stated in the article, but only within the limiting factor of the relativity of the universe to these measurers, these observers' census estimates. A self-propelled, accelerating, fast traveler in the universe, in a more elastic 4-dimensional space-time and mass-energy expansionist-contractionist, inflationary-deflationary, relativity could find himself dealing in a universal average density of 12, or 24, or 48, or..... hydrogen atoms per cubic meter volume (as measurable by him). In other words, he could be turning a certain observation on its head. He could very well be shrinking the universe around him. And instead of dealing in just one observer in his 4-dimensional space-time environmental balloon at one and the same time, he could be dealing in dozens across a shrunken [relative] universe in that same moment in time.
Of course in any deceleration his universe would reverse from a steadily contracting universe to a steadily expanding universe (as he drops from plane of universe to plane of universe). And, likely, from many hydrogen atoms per cubic meter volume back to the very few as stated.
The universe traveler's relative 'local' universe is not the Earth's observers' relative 'local' universe. Separation in space-time results in separation of universes. Relativity will breakdown. What the Earth observer, astronomer, physicist, claims for the distant universe (u), or the Universe (U) at large, may not be what [is] concerning the distant universe (or for the Universe at large). That figure of 31% average for matter just seems to me to be right for an average. My clumping it, producing an analogy to Earth's surface division between land surface and sea surface was just my analogy concerning "surface".