There are many better long term systems than the turtle method. One of the problems is that it keys off intraday data, so you need to call in stop orders for all the markets you trade every day. Most brokers won't like you ringing in with 20-30 orders each day, especially as only a few a week will give fills.
A simple system that does much better than the turtle breakout is as follows:
Compute the mean (m) of the last x day's closing prices, and the standard deviation (s) of the last x day's closing prices.
If today's close > m + s go long on the open tomorrow
If long and today's close < m exit long on the open tommorow
If today's close < m - s go short on the open tomorrow
If short and today's close > m exit short on the open tomorrow
A good length for x is between 60 and 100. You can also use a multiplication of s such as 1.5.
This system is good because it will catch all major trends, and is easy to implement (you just ring the broker the next morning with your market on open orders).
For money management compute risk as (entry price - m). You can also dynamically adjust by considering volatility (ie risk more when volatility is low).
As with all systems, you will get whipped to death when the market is not trending, but in the long run you can expect a run similar to the average CTA.
For portfolio selection, pick one or a few products from each complex (eg currencies, interest rates, stock index, grains, meats, metals, softs, energy).
The only way to long term trade is to be highly diversified, so you really need over $100k to do it properly. As a previous poster said, you make all your money on less than 10% of trades so you need to get on every trend.
