Gotta justify the need for a widening for more than a handful of days a year. Nearly every road needs a widening based on that criteria.
If it was congested multiple times each weekend all summer long, then it's worth looking into a widening.
This is why the starting point for a decision as to how much capacity to add is the number of lanes required to accommodate the design hour volume at a target LOS that is often specified as LOS B for rural freeways and LOS D for urban freeways. (The justification for using an inferior LOS in urban areas is that congestion is more tolerable in small doses.) DHV is traditionally defined as the 30th highest hour in the design year, but in special circumstances (e.g., very high cost to widen and extreme seasonality of traffic) can be something like the 100th highest hour.
On this forum, we tend to discuss whether widenings are justified or not in terms of very crude present-day AADT warrants (e.g., 10,000 VPD in flat country for widening from two-lane to four-lane divided, or 30,000 VPD for freeway widening from four to six lanes) because that is the only form of traffic data that is conveniently available from most state DOTs. The traffic modeling on which most actual widening decisions are based tends to be more closely held.