Because it could lead to an algorithmic discussion of numerical integration by shells vs numerical integration by disks from integral calculus, a discussion about 1-3 sig fig engineering estimation and error estimation and error propagation, maybe a discussion about circle packing efficiency.
It's an analysis of four things: How an applicant socially responds to ridiculous requests outside their previous experience, and how they handle initiating and architecting a "large" engineering project from a vague request, how mathematically well educated an applicant is, and how wide the knowledge level of the applicant extends (basically a psuedo IQ test, with all the legal risks that implies).
This is vitally important for a handful of positions, but a complete waste of time for most positions, thus silly.
It's an analysis of four things: How an applicant socially responds to ridiculous requests outside their previous experience, and how they handle initiating and architecting a "large" engineering project from a vague request, how mathematically well educated an applicant is, and how wide the knowledge level of the applicant extends (basically a psuedo IQ test, with all the legal risks that implies).
This is vitally important for a handful of positions, but a complete waste of time for most positions, thus silly.