Software to analyse GPX running files in-depth

Hello, I need an OS X based tool which can analyze GPX data in-depth. Background: all fitness sites that I am aware of which can take in GPX data give a relatively superficial analysis of runs: - heart-rate over time, av. / max / min HR - location plotted on a map (neat but not very useful) - Speed over time - etc. What I am looking for is a tool which can properly correlate these data in order to derive a fitness score for a particular run. The challenge here is to understand the relationship between Heart-Rate, Speed, Incline / Decline, Duration of the run taking into account body weight, gender, and age. For example, many existing websites would say I did a great job if I completed a given run in 30 rather than 40 minutes. However, they do not take into account the heart-rate that was clocked vs. the time it took on an integrated basis. Specifically, the tool needs to be able to: - import a GPX file, weight, age, gender - provide a graphical display of time (x-axis), Heart-Rate, Elevation, Pace, Grade-adjusted-pace (all y axis) - compute and display the fitness score for a particular run; this score is something you must develop, but it needs to be based on the entire GPX file, not on averaged-out data (i.e. av. heart-rate, av. pace, distance). I have ideas on how I would do the computation but would like you to come up with your own. - this only needs to work for runs, not for biking etc. In order to do this, I am providing a set of GPX data you can work with. You should in addition work with: - weight: 60kg - Age: 35 - Gender: Male The tool does not need to be: - pretty - scalable for many users - able to display the run on a map Projects which are produced based on uploading the GPX files to an existing online fitness website and then using their analysis are not acceptable. This needs to be a stand-alone piece of software. If this project produces valuable insights I will consider pushing it forward into something more serious. If you have questions please ask!