To add a slightly different spin on this fascinating issue, I think there are several key questions.
First question is, are you trying to predict volatility or direction? Second question, how is implied volatility priced? Thirdly, how comfortable are you with stop losses. Assuming you have no access to guaranteed stops are you happy that you will be ? Fourthly, how expensive is to trade the underlying future? Finally how liquid are the options?
If you're trying to predict volatility and you think volatility will rise
AND implied volatility is relatively cheap
AND the underlying is expensive to trade
AND options are illiquid
... then you should buy option straddles or strangles.
If implied vol is expensive, or options illiquid, and the underlying is cheap to trade, then creating a synthetic option like payoff by following a trend following strategy could be better.
(If you think implied vol is expensive, and you think vol will fall, then you will have to sell strangles and delta hedge them. I mention this only briefly as its not possible to capture this with futures, and its relatively advanced and dangerous territory)
If you're trying to predict direction
AND implied volatility is relatively cheap
AND you're uncomfortable that stop losses will protect you, and you want the guarantee of a maximum loss (the premium)
AND options are liquid
... then you should buy calls (or puts)
If implied vol is expensive, or options illiquid, and you are cool with stop losses, then you should buy (or sell) the future and use trailing stop losses to protect yourself.
Note that I don't believe questions of margin should come into it. The trade should be sized so that for the median outcome about the same amount of risk is at stake, and roughly the same margin is required (although there will be variations during the life of the trade). If you're picking your trade based on margin, then you're probably being too aggressive.