The Graham Number (the square root of 22.5 × earnings per share × book value per share) is Benjamin Graham's ceiling on what a defensive investor should pay: 15× earnings and 1.5× book at most, combined. It's deliberately crude, and deliberately harsh on modern asset-light companies. Apple's huge buybacks have shrunk its book value to a few dollars a share, so its Graham Number looks absurd. That's the model showing you its own limits, not a bug.
Earnings Power Value asks a blunter question: what if this business never grows again? Take its normalized earnings (we use the five-year median) and capitalize them at the cost of equity. What you get is a floor — the value of the business as a frozen cash machine. Anything the market pays above EPV is a price for growth, and the gap tells you how much growth you're being asked to fund.
Neither model needs a growth forecast, which makes them useful exactly where forecasting growth is silly: cyclicals, slow compounders, and companies where the DCF's guards fired.