U2415, factory calibration?

Dell "says" that its U monitros are factory calibrated to a mean error less than some value, I think that it's 3 deltaE (no claims about what version: 76, 94 or 00)

It's a mean error, there are values over and under mean value. For example white, you can get a white "color" (255,255,255) far enough from D65 (internet & video stantard white) to be noticieable, but MEAN error keeps under spec.

A sRGB screen could be ouse out of the box if white and grey ramp (under no color managed enviroment) shows a neutral color, aka "no color". If you are fine... it's a sRGB display, no calibration is mandatory, but the only way to test it is to measure it.

With widegamut displays like WG CCFL (U241, U2711) or new GB-LED Dells (U2413.. etc) a proper measurement device IS MANDATORY, with all display manufacturers (dell, nec.. ). Today best measurement device for this kind of photography displays is i1DisplayPro colorimeter. It's more accurate by far that low cost spectrophotometers from Xrite like i1Pro2 (or lower spec'ed brothers).

A colormunki display shares hardware and accuracy with i1Displaypro, BUT:

-it's 5x slower

-you CANNOT use it to internaly callibrate ANY monitor (GB-LED dells, NEC, Eizos..) because Xrite blocked drivers to do that, and those companies rely on colorimeter manufacturer drivers.

-Xrite software for Munki Display *s*u*k*s*, but i1Profiler bundled with i1Displaypro is not as good as ArgyllCMS. My advice with non hardwre calibratabe displays like yours is to use ALWAYS ArgyllCMS and its frontend DispcalGUI.

You have a WLED sRGB IPS display, so you MUST USE WLED spectral correction with calibration software. "Spectral correction" is a way to make accurate the measurements taken with this high performance colorimeters and is independent from actual spectral sensibilities of YOUR unit. "Matrix corrections" does not have this feature, so if posible use "spectral corrections" (EDR o CCSS files in Xrite or ArgyllCMS naming convention)