The actual interest rate isn't really that important: Its stability and predictability is. In practice, every loan and long term contract that is made is implicitly a bet on expected future interest rates. We could write contracts expecting a federal reserve aiming at 2% NGDP growth year to year, or we could do it at 5%, or 10%. There's an effect due to Money Illusion, but if the rate really is predictable, the difference isn't that big.

Now, when people make long term bets, like buying a house on investing in a business, and the expected inflation rate changes significantly, instead of their economic behavior being the key part of the success of the loan for both sides, it's the change in rates. We've seen banks in trouble not because 7% is too much, but because the borrowing decisions they made, expecting 2% inflation, proved to be major losers. Similar things happen in the real estate market: Ordinary borrowers are basically paying double, month-to-month, on the same house bought today than with a mortgage bought 2 years ago. So anyone that built new housing, doing the math to people buying the house for X price, is either seeing less demand, as only people that don't need a loan can afford it, or just a lower sale price, as the same salary can now pay less for a house. Either way, the builder loses on the bet, and houses either don't get built, or become less profitable.

So really, there's devastation either way when inflation or NGDP predictions are far off. And that's how the fed fails: Just not meeting targets in either direction.