I fully supports your proposal

I am also an advocate of multiple contexts, but for an entirely different reason.

I would never ever use multiple contexts in the way you describe, i.e. apply a context tag for each and every kind of situational factor under which it is still POSSIBLE to do the task, e.g. @Phone, @Email, @Sofa, @Car, @Kitchen etc if you write a task to "Remind my wife that I will pick her up at 7 p.m. tomorrow". Instead I would just write down any context that is UNIQUELY NECESSARY. In this case there is no such context limitation, so if I still want to apply a context for list grouping purposes I would put it down as @Info (@Anywhere).

But I think multiple contexts still have their very good uses. Instead of defining a long list of ridiculously awkward variations of my contexts I could have a shorter list and combine them as I go using filters.

For example, with only one context, I have seen many people post lists with maybe hundreds of contexts on forums, such as:

- In the park with Mary, Paul and the dog
- In the park with just Mary and the dog
- In the park with just Paul and the dog
- In the park with just the dog
- In the park with Mary without the dog
- In the shopping mall with Mary and the dog
- In the shopping mall with just the dog
.. and so on for miles and miles ... horrible

If you can apply multiple contexts, the only contexts you need to have defined in order to cover all the above cases is Park, ShioppingMall, Mary, Paul, Dog, and if you apply only the contextual factor that is really REQUIRED (NECESSARY) you will typical have just a rare few contexts applied to each task. Very manageable. For example, if the task cannot be done unless you have both Mary and the dog present you apply the two contexts Mary and Dog. That's it.

Doit already has multiple contexts ;-) The sad thing is that the other kind (called Tags) are invisible in the task list and therefore cannot be trusted. Another sad thing is that the Tag contexts are totally uncoordinated with the Context contexts.

@Folke
To use your example -
For tasks that can be done by a computer only, I put the context @Computer.
For tasks that can be done by a phone only, I put the context @Phone.
There are tasks that can be done by either a computer or a phone. Let's say I put them in the @Info context like you suggested.
Let's say that I sit by my computer and I want to see all the actions that I can do with my computer. I'd have to check the @Computer and @Info contexts to see all the possible actions. What I'm saying is that I'd like to see all the actions in one list, without having to check multiple lists.
So for the actions that can be done by either a computer or phone I'd like to be able to put 2 contexts - @Computer and @Phone, so that the actions in that group would appear both in the @Computer context, and the @Phone context, and I'd be able to see all the possible actions in one list, without having to check multiple contexts to see all the possible actions.

I agree with you that the second way of using contexts that you suggested is also useful, in order to add several contexts that all of them are required to complete the task. The difference is that I'd like the feature to choose 2 contexts - @Computer and @Phone, so that the tasks in that group would appear if I select one context - @Computer OR @Phone, whereas you'd like to be able to select 2 contexts - @Mary AND @Dog and see only the actions that have both of these 2 contexts. I think that both ways of using multiple contexts would be useful.

What do the developers think?