IMO for backtesting options strategies this method is well good enough. This is "backtesting using an options pricing model based on real historical prices of the underlying", not "backtesting using real historical options prices". And since in real life volatility is not constant: one simply can use different volatilities (for example slightly increasing or decreasing, or recalc it at each bar) in the backtesting loop, ie. even that can be tested using the said approach, one just needs to do some programming...
option data is useless and harmful. only idiiots want to pay mony to buy and study. trading ideas on options are created from the underlyings.
Exactly! Historical data of the underlying fed to Black-Scholes is a perfect combo for studying options strategies.
And where would you get the vol surface to plug into the Black-Scholes model? Without the options data?