One problem with how people are thinking about duration: they are thinking of a static one-time change. If bond yields go up once, after waiting years equal to the bond fund's duration, you break even and start to be ahead. But that assumes only one interest rate increase, which has not been true even for last year.
Take Vanguard Total Bond Market for example, with a duration of 6 years. Does anyone believe there will be just 1 interest rate increase for 6 years? Last year the Fed raised rates repeatedly, which tends to flow into bond yields. But set aside multiple increases in one year, and consider one increase every 2 years.
0 years: rates +1%, BND loses -6% value but has +1% yield
2 years: BND has gained +2% over past 2 years, but -4% of loss remains
2 years: another +1% rates, same impact (-6% value, +1% yield)
2 years: now BND has -10% value, but has +2% yield
4 years: BND has gained +6% over 4 years, but -6% of the loss remains
4 years: another +1% rates, same impact (-6% value, +1% yield)
4 years: now BND has -12% value, and +3% yield
6 years: so far bond is still -6% down, and will have +2% yield going forward
6 years: another +1% rates, same impact (-6% value, +1% yield)
6 years: all told, BND is -12% with +3% yield
The key thing to note is although you waited 6 years after the original increase in yields, at no point did the bond fund break even. Layers of increases keep the fund's value below your original purchase price. While seeing exactly +1% yield increases exactly every 2 years is unlikely, it demonstrates the impact of repeated yield increases on a bond fund.
Last year rates were raised repeatedly. In January the stock market took a hit when investors realized the Fed will probably be raising rates more than expected this year. So while I agree duration is the break even point, in the case of the Total Bond Market fund it's unlikely that rates will only go up once in 6 years.