Even as an EV owner I really see this as a problem. Lets say you are away at a summer home, two month mission trip, long vacation and there is a storm and the breaker trips on the car. Imagine coming home to find you have to spend more for a repair than most cars cost?
Then, when the car gets to 5% SoC you get a call (or an email?) from Tesla. You phone (or email) a neighbor or friend and ask them to check on it. I always leave a spare key with a trusted neighbor.
Another possibility is last year I had a bad bicycle accident and if it were not for my helmet I would not be writing. A few months in the hospital suddenly became a possibility. Since my wife has no desire to drive the Tesla would she verify it was plugged in?
If you are in a coma for two months, you have much more serious problems than the car. But if there is someone who checks your messages (your wife?) they'd get the message and be able to deal with it. If you are hospitalized but conscious, you get the message one way or another.
The trade-off is that Tesla uses a battery type with greater energy density by both weight and price, but the "cost" is that it requires more care, and therefore more power, than other battery types (some of which weren't even available when the Roadster was designed). With LiFePO4 the Roadster pack would have needed less care, and could have been designed to endure years (?) without being plugged in. But the car would have cost more and had less range. Trade-offs.
The relatively shorter safe no-plug time is the "price" of getting the most range and keeping the cost down.
BTW and FWIW, I took a long vacation, left the car plugged in in Storage mode, and IIRC the pack was at about 20% when I returned home. This was plenty to do some shopping immediately, and the next morning the pack was at a "full" Standard charge. (I had no reason to charge it immediately, so let it wait for its usual midnight start time.
A related question: How much money do you want to spend to reduce the likelihood of an event which is already extremely unlikely? If a different design would have enabled to car to sit unplugged for a year but would have raised its price by $10,000, would it be worth it to protect the car against the event that you are in a coma for two months AND your circuit-breaker trips AND you have nobody checking messages who can deal with the car when Tesla tries to notify you? I think it's a case of dreaming up extremely unlikely scenarios to justify adding a possibly expensive additional fail-safe. The guy who started all this simply neglected to plug in his car. The lesson for the rest of us: Plug in your car.
But, OTOH, if a simple software update can extend the safe unplugged time without making a different fail route more likely, then I'm all for it.