That's what I am thinking. Although the distance to empty must also use the average mpg along with the fuel level to come up with the calculation. Before this update I always just figured the distance to empty was too high just because the avg miles per gallon was too optimistic.
My Scangauge gives me a MTE number that seems pretty realistic using data from the ECM and as far as I can tell it doesn't use tank fuel level at all. Hopefully Subaru has figured out how to use the same data to come up with a more accurate number.
I had a DTE recall applied on my MY 2018, but I don't know if the recall for select MY '17 is exactly the same or not. What I can tell you is that the recall impacted both the DTE and the fuel gauge level/warning light level on mine.
My first tank from new, I ran it until the low fuel light came on and immediately refilled. It took just a hair over 18 gallons (in a supposedly-18.5 gallon tank). At that point I had around 70-80 miles estimated remaining. Based on tank capacity, there was no way I had enough fuel left to even approach the remaining estimated DTE.
After the recall, the low fuel light comes on after aprox 16 gallons used out of the tank, and usually a remaining DTE of around 50 (although it's hard to tell, as I have a mix of very short trips daily, and longer drives maybe once per week, so my mileage fluctuates from a low of around 22 up to around 29 actual measured MPG.
When I took the car in for the recall, it indicated just under a half a tank remaining. After the update, it was showing about a quarter.
The original recall notice from Subaru specifically mentioned running out of gas while the DTE showed remaining distance. But clearly more than just the DTE part of the system was impacted. If I had to guess, the DTE gauge doesn't keep a "running tally" of fuel used, but instead uses the fuel level sender to approximate how much of the tank is remaining, then multiplies by a time-weighted average mileage to get remaining distance. Otherwise how would it tell what the fuel level is, if you add gas but don't completely fill up? I'm betting that either something changed in the fuel level sender that changed the voltage detected at empty, or an incorrect value was set as the minimum fuel level in the cluster, causing the gauge to not read completely empty when the tank was empty. Then that carried through to the DTE calculation, causing it to over-estimate the remaining range by roughly 10%, based on my observations.
Could be worse. I have a 2006 Triumph motorcycle. The float used to detect fuel level is a little sticky, and doesn't always actually float. When I fill up, it continues to read the old value for ~ 10 miles or so, then it slowly increases to almost-full (it will only read "FULL" if I ride it for a bit after filling to jiggle the float back to the top, then turn the ignition off and back on). Then, as the tank drains, the top 1/2 of the scale drops slower than expected (corresponding to roughly 2/3 of the tank volume) then the bottom half goes very quick. And the DTE indication works off the indicated level, so I can run the tank almost dry, completely fill it, and then on restart it'll tell me I have 20 miles. Then, the fuel level increasing while adding distance messes the whole thing up, and it'll start telling me I have all kinds of range (as in 500-600 miles, when it'll be dry in about 200). The only thing I rely on is the trip odometer and the low fuel light, to determine when I need to stop for gas...