After writing my tirade up there a while back I've given the round robin treatment to a few other task management tools, did some soul searching on how I want things to work vs. how they do, and also added some more tasks to my daily life in the form of a new hobby (Photography.)
So here's what I've learned, and please remember this is a personal experience. I'm going to give my pure opinions and you shouldn't be offended if I speak poorly of a methodology you prefer. These are opinions of methodologies for myself.
Firstly, contexts still aren't perfect for me. I have tried several different context schemes ranging from the incredibly simplistic "why am I doing this" contexts to the sprawling "primary dependency" approach. Where I've settled is sort of something in between. The important thing for me was why I wanted to use contexts. I don't need them. Nobody does. You can make a flat list of tasks and expect to get things done. But, this does impose upon you to review all your tasks to see which are best done where you are, or where you want to be.
So for me it came down to that last statement... where I am vs. where I want to be. I started to realize that I didn't want that question answered by the computer/task system. I decided that I would use both. That I would not fear moving tasks between contexts as I found them in the wrong place over time. So now I have a hybrid set of contexts mainly consisting of real things and real places as well as real people (which surprised me.)
The top of my list, and the contexts I use generally in my first attempt at organizing a task is a list of major tools I use to perform my tasks. So I've got my Work computer, my home computer, my iphone, my ipad, and the web in general. I struggled with the web the most of these things. Oh also I have my camera. Anyway, I struggled with the web the most because I knew I could access the web on most of my devices for most tasks but also, there are some tasks I can't do on my iPad because websites focused too much on mousing or use flash or whatever. But, without giving too much thought to a task that is pretty specifically doing something on some website, I decided to just toss them into "Web." I prefer to use my iPad for these tasks but as I'm doing them, or deciding to do them, if I realize I can't do it on my iPad or iPhone or whatever is convenient, I'll make a change! I'll move the task to iMac or MacBook Pro depending on whether it's a personal or work task. I don't like to do my personal web browsing on my work machine.
That was one of my biggest challenges, overcome, basically by accepting fallibility and agreeing to myself that I will review tasks over time, and rather than wait until the last second to see what I should be doing, I'd use the hotlist and always plan at least a little ahead, or otherwise reduce my expectations about any potentially doable tasks (namely, that even if I'm in the right place, and I have tasks, they might not be doable because they might be organized wrongly.)
Furthermore, I found a lot of tasks really want to be in a pretty specific place. I almost always end up with a new place for these types of tasks. An example is a plan to go take pictures of a specific area, or of an event in that area. So what I've decided was to relax, and create lots of these contexts, then just organize them under a catchall so they don't cause me problems when I'm planning or reviewing. The hope here is that over time I'll find the locations that are frequented, and iteratively build in some entropy to those locations when I'm deciding what context to put a task in. So again, this relies on review and iterative changes on the fly over time. It's working great so far though.
The last set of contexts is people which I always struggled with the most. The reason it fouled me up was that I felt like people were the thing I had the least control over. I hate having tasks that I can't get done because of other people and feel I should really just be giving them the task and waiting on whatever next task I have that depends on them. However, once I gave on perfection and started tossing in contexts all willy nilly, I found comfort in certain tasks, especially research related tasks, being assigned to people contexts. Also, I have a few people that basically need to be babysat by me so they get tasks in their context that are like... "make sure dude did the thing" and on those I will do the due date + repeat thing so that I remember to bug them weekly or whatnot.
So what it comes down to is basically this: Different projects and tasks, different types of things I want to remember to do in different areas of my life begged for different meanings of the concept of contexts. I don't only use them to decide what I can do but what I want to do. I rely on reviews more than ever, and am so glad OF for iPad has a great review interface because of this. Under certain circumstances, such as when I know I can't leave the office until 6pm or something, I let the contexts inform me of what I can be doing. In other circumstances, such as when I'm awake earlier than usual on the weekend and have nothing to do, I sit down, relax, run my review and look at what things I may want to do. I even have a conceptual context dedicated to the things that I can't justify a priority for but would happily adjust my day a little to accommodate when I'm bored, such as watching a movie or going to a nice dinner.
Tossing together a variety of uses of contexts is only working for me because I've accepted that my daily, and weekly routine needs to include reviews. I also adopted the hotlist concept using flags, though I always push myself to set due dates instead if something is important and not simply something I wish to get done.
So each day I'll quickly review available tasks, look over the contexts where I know I have to go, look over some of the projects that I've been wanting to approach to decide on additional places or people I might need to track down, and coupled with a "Today" perspective, I end up with a list of stuff to get done, or at least try that works out pretty well.
Making constant adjustments to due dates, flag state, contexts (including adding new contexts and recombining them over time) has made contexts very useful to me despite the reality of my tasks being done mostly in a couple of places.
Oh right, one other thing, adding location contexts freely has given me much greater value from iPhone OF's location feature, and will likely derive even more usefulness out of iPad OF's added mapping capabilities. See, even though I might have 25 different location contexts, and a lot of them pretty close to one another... well looking at one of those contexts isn't going to inform me of what's happening nearby (or rather, what could be happening) but once I pull up the map on my iPhone I suddenly get a listing of all the nearby contexts. So what once was an average 2:1 task:context ratio is suddenly a nice comprehensive list of tasks that I might perform where I happen to be (or on the iPad, where I plan to be).
So my previous opinion about the location capabilities of the iOS OF releases is completely negated now. It's become a vital way to coherently aggregate contexts in a meaningful, spatial manner (for the contexts where that makes sense).
Interestingly enough, I now feel that the desktop OF is lacking in some ways for the way I've adapted my usage of OF. I get a dramatically more effective experience from the iPad version for planning and the iPhone version for momentary status checking and recording. Being able to add a context then go set the current location of it on the fly, for example. So I hope to see the improved review and forecasting interfaces as well as an appropriate location interface added to the desktop OF in upcoming versions.
I hope that wasn't too difficult to follow. Suffice to say, I'm much happier with OF now that I've tried a few ways. I also wanted to make a couple of points about the competing products I tried:
1) I didn't find it any easier to manage multiple tags in lieu of contexts. What I found was information overload, and a compulsion to tag things with what ended up being meaningless details. I found this much harder to reconcile later. Also, with, say Things, I couldn't find many ways to make tags useful. I could associate them with locations or other real life waypoints. They basically became search terms which were harder to use than just plain searches.
2) I tried toodledo coupled with Tasks apps for iPhone and iPad and I really liked the improved repeating task capabilities (I could do last friday of the month, which I can't do in OF) but still found it generally too limited. The experience only further emphasized my appreciation for my own determination of how deeply I should nest tasks and outline projects.
Now that iPad OF is out and the added features there... I really doubt I'll be switching tools ever again (assuming nothing breaks or the world doesn't end, etc.)
Though I'll still be adapting my various context/project planning strategies, I think I've come to a conclusion about what to do with contexts when the primary colors approach that is often exemplified doesn't work and it's simply a matter of relying more on the other aspect of OF/GTD that is lesser advertised: Review, Review, Review. Don't fear adjustment, and don't sweat the idea that you might've gone the wrong way with a particular context decision because you can always change it later! Couple this with always using due dates on anything remotely important and you'll never miss a task (over time) and I think (over time as well) you'll (and by you'll I mean I'll) improve the success of the tool's effectiveness in helping to actually accomplish your goals with less stress, and better timeliness, etc. etc.
Bye.