The reason for 1M trading for the backtesting is that results of trading 1M should be statistically more significant than trading smaller amounts. Ie. it is usually harder to make higher profits with big money than with small money. The more the capital the less the % performance and vice versa. If the system masters this then it will do usually better for small money. Getting high results for big money is a challenge because one has to take some more factors into consideration like max position size, to be realistic.