It's raining today so I did a calorie test using my truck, no HR of course, and the 705 calc'd calories with just speed and distance and gain. I recorded one ride driving up a long steady hill and then a second separate ride back down. Managed to start and stop at almost the same spots getting the same distance and kept the same average speed both ways.
I was very impressed with the results. Uphill 585 calories. Downhill 8 calories. Obviously the gain is a major factor in the calc's because that was the only piece of data that changed. Here are the files. You can open them in TC and see for yourself.
There was one major error. The total gain calculated by the 705 was 145ft on both directions whereas the actual elevation difference shown on both profiles was 240ft and Google Earth shows 240ft of elevation difference also. That is a huge software error. The recorded trackpoints and profile agree with each other and Google but the 705 software calculated the wrong total gain for both rides.
Is this a software bug?