Alternators are well known for sapping about 10% of a vehicle's fuel economy and that's assuming a fully charged battery. Don't believe me? Go on any hypermiling forum and one of the things discussed was removal of the alternator belt and just charging the lead acid battery when you reach your destination.
The primary resistance from the alternator is resistance from the belt itself and from the inertia or the pulleys and resistance from the bearings. Removing the alternator entirely is going to make a crap-ton more difference than just changing the alternator duty cycle. The fact that you think removing the alternator entirely is comparable to changing the duty cycle is a perfect illustration of why you're mistaken here.
Let's look at a typical alternator. You're dealing with a maximum output of somewhere around 130 amps. At 12 volts, that's 1,560 watts. For simplicity's sake we'll call it 1,600 watts.
If we assume a 50% efficient alternator, that's 3,200 watts of power sapped from the engine (not counting frictional losses from the belt and pulley system). This is 3,200 Joules/second or 11.52 MJ/hour. Gasoline has 120 MJ/US Gal.
With a 20% efficient engine, a maxed-out alternator will use 11.52/(120*.2) gallons of fuel/hour, or 0.48 gallons/hour.
That's the difference between zero output and max output, so it's easily double the difference between normal load and full load, so we're dealing with maybe 0.24 gallons/hour of difference in fuel consumption.
OP's vehicle was rated, under the old, optimistic EPA standards, at 23 mpg highway. At 60 mph and 23 mpg, that's approximately 2.6 gallons/hour. Adding another 0.24 gallons/hour brings us to 2.84 gallons/hour and brings mileage down to 21.12 mpg. For a difference of 8% (1.88/23).
I was off on my initial estimates of the effect of the alternator being run at 100%, but that doesn't change the fact that you're just plain wrong if you think that the main culprit here is the battery charging. Even if he got that 8% improvement over his current 17.7 mpg back, he'd still only be up to 19.1 mpg (17.7*1.08)
The OP's vehicle is EPA rated (using the newer, more accurate numbers) at 17 mpg city. The OP is driving only 3 miles, still in warm-up enrichment, in cold weather. That's going to be the biggest issue for him. He could remove the alternator entirely he still wouldn't get back to the 20.9 mpg he was getting before.
The major player here is simply a change in driving habits.
As for OP "operating in open loop", unless his car was manufactured over 20 years ago, most cars enter into closed loop within 10 seconds of start up... that's what oxygen heaters are for... (cars that don't have heated oxygen sensors take significantly longer to enter into closed loop operation.)
OK, you caught me, I played fast and loose with the terminology. It will still be in cold-start enrichment. Your terminological nit-pick has no practical effect. The fact remains that the vehicle, with a cold engine, will still be operating with a richer mixture even if it's technically in closed loop.
ZV