This is a great example of how selective use of statistical data especially "averages" can be used to try and sell a false premise.
Picture a conceptual graph using average hand-picked stats like pts, rebs, assists, etc. of LeBron's first 4-5 years which weren't nearly as "good" (averaged 20.9 pts his first year) as his next 6 to 8 years with this year being "around average". So even if this year is "average" in the hand-picked stats, LeBron can be shown to be performing well under his peak years.
Moreover, moving to advanced stats, his W/48 win share per 48 minutes this year is 0.187 well below all but his rookie year of 0.078 (high was 2012-13 at 0.322). His VORP (Value Over Replacement Player ) is 4.0, again lowest except for his rookie year of 3.1 (high was 2008-9 at 11.6). His PER (Player efficiency Rating) this year is 25.7, the 3rd worst year to year 1 and year 4 (high was 2008-9 at 31.7).
So to my eye, he obviously is still performing at a high level, but well below his peak and pretty close to a bell curve shape decline so far.