Agreed, not totally useless, but there is a long period in your graph where not much is happening
Be nice if these were winning trades. Perhaps it needs the famous 'BatchBoy' optimisation. Mind you - it doesn't follow the 'scientific method' so is not repeatable in live.
If there is enough demand I might make the opening/closing of trades occur on the boundary of a 1 minute bar. That would make testing a bit quicker too.