Same can be said for SFO, just replace "thunderstorm" with "fog/low ceiling/too-close-for-comfort parallel runways".

Of all the routes that UA does do some serious padding on, JFK<>SFO is probably the most justified. ORD<>LGA is a close second.

Also, in fairness, if you compare the block times to the actual flight times of routes that aren't as susceptible to weather/ATC delays, they are usually much closer together. DEN-MEM is one that I take just about weekly. I can't recall the flight ever departing or arriving more than +/- 10 minutes from scheduled, with the lone exception being a recent 3+ hour delay due to weather across the top of the country from Chicago to the east coast.

I have noticed this occurs during peak travel times at major airports. I think it is partially to allow for long waits on the taxiway for takeoff. If you are number twenty for takeoff, you want that extra time, especially if you have to make a connection. As for arriving early, often your gate is occupied and you have to wait anyway.

Padding is not new.
 

This has been going on for some time now. It's great. I am always early. I really like how they do it. It's the old motto: "Under promise. Over deliver."

It's working well for UA, in my book. ^

True, but on United's p.s. service (the original topic of discussion), there's probably room for a lot more padding. You're not going to get an extra turn out of the aircraft by decreasing block time, nor are you not going to get your 2 (maybe 3?) segments per day by increasing block time.

It's probably more a factor in aircraft running short segments, and I'm sure the scheduling system can figure out how much padding they can get away with before it starts to affect aircraft utilization.