The paper studies a way of comparing finite trees where each node carries a label drawn from some well-quasi-order (a set with a "nice" ordering that prevents infinitely long decreasing sequences or infinitely wide antichains). The comparison method, called the h-preorder, says one labeled tree is at most another if you can find a structure-preserving map from the first into the second that respects both the tree shape and the labels. Understanding how complex this ordering becomes on large collections of trees is the central goal.
Two key measurements capture that complexity. The "width" of an ordering tells you how many elements you can find that are mutually incomparable, meaning no one of them is below another. The "maximal order type" is a more subtle number that measures, in a precise set-theoretic sense, how long the longest well-founded chains and interleaved structures can get. Both quantities can be assigned not just to simple sets of numbers but to the richer orderings that arise from tree comparisons, and computing them is generally difficult.
The authors prove exact, tight formulas expressing the width and maximal order type of the h-preorder on labeled trees directly in terms of the corresponding width and maximal order type of the underlying label ordering Q. In other words, if you know how complicated the label set is by these two measures, you can precisely predict how complicated the resulting tree ordering will be, with no slack in the bounds. This is a sharp result in the mathematical study of well-quasi-orders, an area that connects logic, combinatorics, and theoretical computer science, particularly in questions about when algorithmic problems over structured objects remain decidable or tractable.