The costs are quoted separately because they are charged separately, and each one sounds negligible on its own. That framing is the problem: nobody pays one of them.
The full list for one round trip
Fees, twice, at whatever maker or taker tier applies. Spread, twice — and on an options position quoted in volatility, that spread is larger in premium terms than it looks. Impact, twice, scaling with size against available depth.
For anything held on a perpetual, funding at every settlement interval, which is a cost of time rather than of trading and is the one most often left out entirely. On-chain, gas on every transaction including the ones that revert.
The asymmetry that ruins estimates
Entry and exit are not symmetric. You choose when to enter and can wait for a good fill. Exits are frequently forced — by a stop, by a thesis breaking, by margin — and a forced exit crosses the spread, in whatever depth exists at that moment.
So estimating a round trip as twice the entry cost is systematically optimistic, and it is optimistic by the most in exactly the conditions where you will need the exit.
Where it bites hardest
Cheap-looking instruments. A far out-of-the-money option with a low premium can carry a round-trip cost that is a large share of the premium itself, protected only partially by fee caps. A thin on-chain token can cost more to leave than the move you were trading.
In both cases the low absolute price reads as low risk, and the percentage cost is the highest on the board.
The habit
Compute the round trip at your size, at fills rather than mids, before deciding whether an idea is worth trading. An edge that survives on paper and not after costs is not a small problem — it is the strategy being wrong.