Thanks for the post and in response to your issues:
To calculate the current weightings of the portfolio the program multiples the final observation price by the number of units for each investment. If return data is used, then only the unit value is used.
The optimization process returns the optimal Sharpe ratio for the entire portfolio (return to risk). Since this takes into account all investments in the portfolio and the correlation between them, the result will differ from that of the individual investments in the portfolio. In the case that you detail, it would indicate that a higher weighting should be allocated to the investment with high return and low standard deviation; however this will also be dependent on any constraints entered and the relationships with other investments in the portfolio as mentioned above.
