It looks like prior link no longer works, try the following.

(PDF) Earned value method as a tool for project control

Earned value method as a tool for project control

6. Conclusions

……. Many organizations worldwide adopted Earned Value as a standard management tool (e.g. US Department of Defense [13], an Australian standard [14]). It is described in practically all management handbooks and incorporated into management software. However, if to be implemented, the method should be used according to its purpose: it is not a tool for forecasting; instead, it facilitates progress monitoring, determination of project status (on time? to budget?), identification of potentially negative occurrences and a rough estimate of their combined effect on the project’s outcome. If the project is to be managed consciously, these occurrences should be then investigated into by means of more accurate methods.

SDPM Approach - SDPM Approach to Project Planning and Performance Analysis.pdf

Slide #50:

Earned Value Analysis is another method that is used for estimating program/project performance.  But this method shall be used very carefully and only in combination with other methods because:
• the real situation may be distorted,
• project managers are motivated to do expensive jobs ASAP and cheap jobs ALAP,
• it does not consider if activities that were performed were critical or not,
• it does not consider project risks.

Slide #57

Earned Value Analysis supplies management with the useful information on project status.

•But it shall be used carefully and only as the supplement to other methods of project performance measurement.  In any case it is useful to analyze EV parameters trends … Positive trends show that recent project performance is better than expected. Negative trends show that buffers are consumed faster than expected and corrective actions may be necessary.

Contractors-float

For most construction jobs we are not required to use probabilistic methods, in this case instead of tracking probability trends we track deterministic trends for optimistic targets we feel comfortable.  We target for early completion in the hope as the unexpected happens we met our contractual obligation, we target to keep some Terminal Float.  In this way we keep it simple, we speak a language everyone understands.

If you want a prediction of likely completion the two methods that work are:

1. Earned Schedule (ES)

2. Earned duration

Both are based on significant data that shows consistently the best predicter of future outcomes is performance to date.  Project manager's don;t like this - their project is always special; but data from 1000s of projects tends to suggest they are not!  Traditional SPI cannot predict time outcomes (it is a cost measure). 

My prefernace is Eaned Schdule over Earned Duation simpl because it fits into an overall EVMS with minimal additional work.  The answer from ES is not set in concrete - its simply the most likely outcome; to change the outcome you actually need to change the way work is accomplished, not analyse probabililities. Management needs to do things to improve work practices, improve resource availability, reduce interfearance in the work, etc.  Then at the next update you can see if the implemented changes have been effective, if that are, keep doing them, if not try something different. 

For more on this see: https://mosaicprojects.com.au/PMKI-SCH-010.php#Process6