Looks like the 5566 has more distance between the rows compared to the 87427. Just a quick look it appears the 87xxx has the same spacing in all directions.
Many connectors are made with all kind of subtle differences which need different footprints. There can be as many as 100 different footprints for a āsimpleā micro-USB connector. Some connectors are made in a pitch of 5mm and 5.08mm, and you only start noticing the difference if you use that connector with more then 4 pins.
Connectors are probably the most cumbersome and error-prone parts for PCB design. Second place is for SOT-23. This is a particularly troublesome part because the pin numbering is not standardized. Itās also one of the oldest small SMT footprints. I assume that drawings of housings got faxed around the world back then and some of those faxes were unreadable. https://duckduckgo.com/?q=sot23&t=hd&va=u&iax=images&ia=images
Let me just throw in mistakes are made because of copious experience also. A well practiced and experienced member of the Arduino forum admits to having a closet full of ācoastersā because he decided not to order a prototype first because it was a āsimple boardā and he was sure it was right.
One post that sticks with me from the EEVBLOG site is a pro that said he does 100% of his own footprints because he canāt afford for them to be wrong. For those of us doing simple projects we get the parts and print out the board to check sizes before ordering the board.
They are a case in point for my philosophy of life and engineering, which states that everything is more complicated than you think. Besides the example of 5.0 vs 5.08 mm pin spacing with which you might be able to mount the wrong one if the connector has only 2-3 pins: I once designed a battery charger with a circular DIN output connector (I think it was 5 pins). In theory this was a polarized connector which could be plugged in only one way. But with persistence and moderate force a user could momentarily mis-connect 2-3 pins of the mating the male and female, causing my carefully-designed charger to go POOOOOOF!
If I remember correctly, we Americans started this Integrated Circuit nonsense with DIP pins on 0.1 inch centers, but then the Europeans went and confused matters by introducing metric spacing. I am looking at you, Philips.
Then came automobile tires many of which sizes combine metric and inches.
Yeah, where connectors are concerned, buy them first then measure and design the board. I got some right angle headers and sockets for mating two boards. I thought 0.1" pitch so simples, no? I forgot the Z axis. The headers were 0.1" off the board, while the sockets were 0.2". I kludged it by soldering the headers 0.1" above the board.
I generally think this is a bit of an overreaction. At least the official library is likely to be better than what a single person can achieve as every contribution is double checked (It is very rare that i get a contribution where i do not need to request some changes).
Every footprint is checked against exactly one data sheet entry and therefore guaranteed to be valid for this one order number. This must always be kept in mind. So all guarantees vanish if the order numbers do not agree fully.
In more detail:
The footprint description states that this part is valid for the following order numbers 5566-24A, example for new mpn: 39-28-x24x As per datasheet http://www.molex.com/pdm_docs/sd/039281043_sd.pdf
Your component has part number 0874272442 (according to digikey) which is well not covered at all bei the pdf and also not by the part numbers listed in the description. This seems to be a newer part as molex does not give an āEngineeringOldPNā entry on its website. Which in this case means that the series number does again fit with the part number (which is not the case for the old connectors).