I was a very late adopter of Glocks during the Gen 3 era, and pretty well settled into Gen 3 9mms (as in tracked new Gens but never moved to them) backed up with a good stock of supporting spares parts, and installation of the now unavailable factory "-" connectors (because Glock trigger feel is their worst point) and factory extended slide release levers.
I pretty much sat out the 9mm Gen 4. The primary benefit to going with Gen 4 was the G23/.40 S&W models where they got the recoil spring assembly correctly engineered for the caliber versus Gen 3.
Now that they are finally phasing out the Gen 3, if I was starting today my first purchase would be the G19 Gen 5 MOS. Going with the MOS provides the option if they finally perfect a durable pistol mounted red dot, the ACRO Aimpoint is getting close if they can resolve some first gen issues. But the main reason to get the MOS model is it does not have the large idiotic semi-circle cut out at the bottom front of the grip. It's my recommendation now to friends and new shooters.
With all due respect and many thanks to Tj for hosting this subguns refugee camp, bashing Glock versus SIG does not track in any way with real world experience. Glock has had problems with specific models or generations or parts changes, and these often shake out with the large sample sizes in those police departments where they actually get out and shoot and train. In the 1990s if you asked me to pick any firearm to pull out of the factory box, load, and step into the next room shooting to save my life I would have asked for a SIG P226. But that ability to trust SIG quality implicitly has not reflected reality for almost 15 years now as SIG QC slipped and became more hit or miss. Very happy with the P220 and P226/P226R, all mine born 2004 or earlier.
SIG modular design and ability to swap grip frames is cool. But you'd think a new US issue pistol would command their full design and manufacturing engineering attention; for business reasons to retain a high value contract, for business reasons of public relations, and for moral reasons that they are arming people going into combat. Yet there have been non-trivial performance and mfg issues with the issue M17.
Glock has individual models that outsell entire product lines or the entire production of other companies per year. After a Glock model has been stable for a couple years it's had a pretty good shakedown of production volume and statistical feedback. Glocks are not perfect, but if you ask me to play the odds including all the unknowns and variables of any -current production- SIG 9mm versus an established Glock 9mm model I'm going to select the Glock. Glocks aren't finely crafted and have a lot of parts that employ inexpensive mfg processes, but they are proven on average against large production numbers to be robust and durable. Replace the recoil spring and the trigger spring every 5000 rounds in any G19 that has proven to extract/eject reliably (one of the previous issues they've had) or selected outside researched model and S/N ranges that had issues, and little else is likely to go wrong. And yes, there is a big Glock not equals "perfection" caveat in the middle of that sentence which is tracking with reality not blind brand loyalty.
Various cost and consolidation issues probably played a role too, but the Navy SEALs replaced the SIG P226R and P239 with Glock 19 variants a few years ago. I don't think they'd be taking a step backwards in reliability with that selection. They'll shoot that one pistol in one pre-deployment training cycle more than most of us shoot any single pistol over several years. And that selection came after several years of US Army Special Forces and other SOCOM use of Glocks in 9mm and .40 S&W during GWOT. So elevating polymer SIG 9mms over Glock 9mms as to quality defined by consistency and reliability can only be accomplished via selective perception.