My feeling is that it's sort of a sliding scale based on the amount of time they have to put in, the amount of credit they get, and the relevance of the task to understanding course material.
It's not uncommon for professors to allow (or even require) students to act as participants in a study (e.g., fill out a survey, be part of an experiment) for credit, on the theory that such participation helps them to "understand the research process". If the tasks they're doing have some relevance of that sort, I think you have an easier case. If it's totally mindless work with no connection to the class, it's more dubious.
Also, at least at my school, every such opportunity must (by human-subjects rules) have an alternative credit opportunity that takes roughly equal time but doesn't require such participation (e.g., write a paper). This kind of alternative is designed to ensure that students aren't forced to work for the professor's benefit in order to improve their grade.
Also, assuming by "50", "60", "70", you're referring to their overall course percentages, that seems like a massive amount of credit to me. When I've given or received extra credit, it's usually been much less than that -- equivalent to maybe 2% or at most 5% of the overall grade. The intent is not to allow students with a flat D to move to a C, but to allow students who have a high D to move to a C. I think offering extra credit that allows students to raise their grade by an entire letter sets some dangerous precedents, especially when combined with the mindless-task aspect.
In the same vein, extra credit assignments usually were the work-time equivalent of say, one homework problem, or at most one homework assignment, expected to take the students maybe 3-5 hours tops, and often an hour or less. 10-15 hours of mindless work sounds like a pretty awful prospect to me. I think it's a bad idea to misuse the extra-credit leverage to have students slaving away for hours and hours.
So basically, I think it is possibly defensible, but more so if the tasks are not truly mindless but have some reasonable connection to the class. Also, I think the amounts of time and credit you suggest are a bit high, and especially so if the task is just grunt work.
Incidentally, as an example, I once was a TA for a class where the professor assigned homework in which the students had to take a spreadsheet and perform certain category-coding tasks on the data. I suspect (but do not know for sure) that the professor was using the coded data for his own research. However, the data was relevant to the class topic, and the coding task, although not exactly an intellectual challenge, was a realistic encounter with this sort of data, in that if students were to later write a paper using such data, they might well have to perform such a task as part of the project. Also, the amount of data coded was rather small (about 50 spreadsheet rows per student, as I recall). My own opinion was that, although such an assignment was perhaps not the best way to get students interested in the class, or enhance their understanding of the material, it wasn't unethical, because it was small in scale and legitimately (if uninspiringly) relevant to the topic of the class.