Some of the 100w MV lamps I have actually run up to over 130v (one goes up to 135v) while others as low as 115v, so I should say it varies between .75a and .80a. I just tested a lamp and the reading was 122v @ ,80a. The 70w ballast OCV is about 250v. I measure current with my Fluke 79 III True RMS. Now I realize that even a true RMS won't be perfect accuracy with HID lamp currents, but I have found over time with many tests and comparisons, etc, that it is reasonably close.
Interesting to add, with these same tests, I have found some HX-NPF 100w MV ballasts to run 100w MV lamps as high as 1.15a....same as the 100w MH ballasts! Another 100w MV ballast I just received runs lamps at .85a, so quite a variation.
I think what you are describing is actual component variation (both lamp and ballast) and it show the immunity of MV technology, but you should count for the fact, then even the replacement ballast will have it's own parameter tolerance, so you should match them by their "ideal" parameters. And if you look into some any more detailed lamp catalogue, you will find, then the "MV100W" need 0.95A, so you need ballast, that will nominally deliver about 0.95A into that 120..130V lamp. Using "70WMH" yield to current tolerance ~0.75..1.1A. If you use "100WMH" ballast, the resulting tolerance area would shift 25% higher, so 0.95..1.4A range, while the upper limit 1.4 would be too much for MV100W.
Btw: sometimes the difference between nominal and actual input power is made on purpose. If you have lamp with "100W" written on it, it does only mean, such lamp shall be operated on "100W" ballast for that lamp category, but actual lamp input power on nominal ballast often differ a lot from it's nominal value, very often on purpose (e.g. someone want to get higher lumens for "M70W", so actually made the lamp as 75Watter by increasing the arc voltage).
