This article, although interesting, is one person's point of view. I would argue that the problem with applying a trend model to S&P is that it's an aggregated index of a large number of stocks. Trend systems seem to work best with "real" currency pairs, e.g. EUR/USD, not GBP/CHF, and sometimes interest rates.
Whatever the reason, it's important to backtest yourself across different markets in order to gain confidence in the model.
The essence of breakout trading is that as a price approaches a new extreme, it will be unstable. It might retrace (it usually does) or it could breakout, in which case you could see a large move. Thus breakout strategies will enter in a small amount on a breakout with a tight stop. If the market continues to move in the same direction, add to the position. It's a bit like creating a synthetic option. In the end, you'll have 60-80 pct of small losses, with 5-10 pct of big winners and the rest small to medium winners.