A few points to add...
If more people are applying each year and the number of available positions remains constant or adjusts at an inferior rate, then competition, on a superficial level, will increase. This is about the only thing someone can say with any kind of certainty.
These numbers are 'averages'. An average number will tell you very little, if anything, about the threshold value needed for matriculation. Individual schools have extremely small sample sizes in the great scheme of statistical analysis. People should be less concerned over the averages and more concerned over a minimum value, if one is present at all. If the minimum values are changing, this should grab a prospective applicant's attention.
As long as an applicant is at or above the lowest value(s) each year (a figure that is not published) - then the applicant is competitive for admissions. Why? Because some candidate has established the precedent or the school has established a minimum value. Most of you will read this and say - well, no one is going to shoot for the bare minimum - that is not wise, nor does it represent a competitive portrait. And that right there is the fallacy of this entire discusion. From an individual standpoint, there is no reason to assume that someone with a 4.0 was any more/less competitive than someone with a 3.0. No one has all of the information at play.
If 5 applicants make up the incoming class and the oGPA values are 4.0, 4.0, 3.0, 3.0, and 3.0, respectively - the average is a 3.4 oGPA. Does this mean an oGPA of 3.4 is the threshold value needed for matriculation? No. Three people got in with a 3.0. Averages don't tell the whole story!
If the following year the class makeup consists of 4.0, 4.0, 4.0, 3.0, 3.0 with an oGPA value of 3.6 - is the school more competitive? Has the admissions department made administrative changes in favor of stronger numbers? Not necessarily. And probably, no.
Are the students that earned a 4.0 more competitive? Again, not necessarily. The admissions process is subjective and multi-dimensional. GPA and DAT values are just a piece of the puzzle. And GPA values are incredibly misleading. Not every GPA is the same and, more importantly, not every institution in which that GPA was earned is the same.
The DAT is certainly more objective in nature - but these are still averages. An outsider has no real semblance over whether the student with a below average AA was any more/less competitive than someone with a 25 AA. And the threshold value is still important here as scores well above the minimum for matriculation can and will influence averages without influencing admissions standards. There have also been systemic changes here as well. Not every test is the same. And the implementation of structural changes like the calculator feature on QR may influence AA scores. The price elasticity of demand for the DAT is also important. If the cost of the test was $25, or free, instead of ~$250 - more people would demand the test and the averages would adjust. This fact can be seen to a certain degree with SAT scores over the past few years as access has expanded throughout the country.
Although it may seem logical and intuitive that higher numbers would make a candidate more competitive - this line of reasoning is resting on a foundation of speculation and assumption. And given the fact that school sizes are so small (~50 - ~100 seats) fluctuations can very well be just a normal function of the process and indicate very little with regards to changes in competition.