...over the ten year warranty period you'll save ~$3000...
That number is way off. You have performed the calculation incorrectly because you omitted the opportunity cost. If you do the calculation correctly, then I estimate (using amateurish cost-accounting methods below) that you end up at around break-even if you buy the battery. Which is extremely interesting, because, assuming rational consumers, as manufacturing costs fall, that would be the consumer price threshold at which these batteries could be marketed and they just started being marketed at that price.
So about the opportunity cost: I have no idea what value is conventional to use as the opportunity cost. But the stock market seems like a reasonable choice. So if you invest the initial $3,500.00 in the stock market instead of buying a battery, and we assume 6.5% returns compounded annually, that works out to to an increase of $3069.98 over ten years. That is the about the same gain., ~$3,000, which estimated in savings on your electric bill after the recouping the battery investment. So you are about $3,000 better off after ten years either if you invest a sum of $3,500.00 in the stock market or if you invest the same sum in a home battery.
Of course there would be more things to consider in making an accurate calculation: differing tax treatments, risk and uncertainty between the choices, what residual worth your battery has after 10 years and its rate of depreciation, the inflation-reduced value of the initial $3,500.00 had you invested it. Also, to do the comparison correctly, assume that the accumulating savings on your electric bill are invested after the battery is paid off. Maybe someone who actually knows how to do cost accounting will chime in here with better estimates. Nonetheless, the claim that you neglected to account for significant opportunity cost stands, even if I failed to calculate that cost conventionally or sufficiently accurately.