Ok, since i meanwhile also got some pms on the topic, here´s a brief (haha) rundown on the different tech from my longtime flash developer view:

First of all when its about some new promising tech coming up while one used something else for ages (after all in my case i´m with Flash since V3) there´s the question: Why should i care about that other stuff, i have done well with what i use for ages, well,for ages.

To me it was like that, several other technologies came, some passed by again and there was never enough motivation to invest as much time into something else as i did with flash.
I learnt to use php,xml,html,js,mysql,zinc and other wrappers, all sorts of graphic apps and what not all else but mostly for doing web apps,screensavers,commercials, games etc, in most cases flash related.
Well, then things changed slowly though, i wasn´t happy with some decissions Macromedia/Adobe made and then also with some tools and flash/flash player versions they released.
There´s way more to it but to just give a few examples in a nutshell:
-I didn´t like it how the policy seemed to change towards flash ide not beeing the one tool for everything flash platform related anymore,instead there were other things released like flex builder, so suddenly flash ide should be the designer tool with no propper ondevelopment of coding related features and for everything else flex builder or some other tool should be used.
-Graphical Editing features weren´t propperly developed on anymore, mayn new features seemed to be only propperly usable with code.
-The performance of the flash player when rendering art or handling code was never raised to the degree i´d like and expected and any raises achieved seemed to be only useful in very specific use cases or go hand in hand with lower level coding manners or come with ridiculously high hardware requirements in return
-I´m fine with coding oop and/or in lower level ways,also things like strict datatyping and what not all if its fitting for a project,its scope and budget but it seemed weird to me to HAVE to do things in certain way in flash purely to get any halfway reasonable performance when doing certain things,especially when if one codes things in similar fashion/language as in AS3 with another language/tool/technology performance is exorbitantly higher than in flash.

The last point is probably the biggest reason that made me think about giving other technologies a more serious look.
Funny, when one thinks about the point that Adobe/Macromedia probably introduced AS3 with the thought that they could attract more developers coming from contender technologies (and this also happens several cases surely). Its nice one can more seamlessly switch from one language technology to the other,but yes,just for me it also makes me look at the others and think when those could maybe be a better choice.

Ok, this brings us to the actual views on the other two,again, from my view after trying both just for a short while, less than two months combined,there´s surely way more to them but i think it should give you a good first idea:

So yeah,i then tried XNA for a while.
For those not knowing XNA its basically the short term for various things:
-An enhancment to Visual Studio, the Coding tool made by MS of which there´s a free smaller version called Visual Studio Express.
-A framework of classes made specifically with focus on game development. There are various classes handling common tasks done in a game scenario.
-A content/asset handling side,processors and importers, basically things you need to get your assets into your game and handle them as needed
-An optional subscription system that allows to deploy games for testing on Xbox 360 directly (called the creator´s club membership)
There´s more to it but i think those are the most important parts for us for now.

So how does it work?
Let´s say we download a feee copy of Visual Studio Express, Download XNA Game Studio,install all,maybe download a few example projects,too (some already come with it).
Now we can start a new project or open an existing one and code things,import assets,test and deploy our game.

Coding is done in C# (in Windows deploy one can also use dlls and other stuff made in other languages,this is not possible in Zune/Xbox 360 deploy since those only run managed C# code).
What is C# like you may ask? Well, it is very similar in syntax and setup to AS3, sure there are various differences but overall its an oop based language,with classes,methods,variables,data types,the common things.
Although its in several ways more strict than AS3 it was actually not that time intensive for me to get going coding something, mainly because the coding tool and debugger are much better than anything available for flash development.

So let´s say we coded our game,what can we do with it now?
There are several platforms onto which one can deploy XNA games,namely right now Windows powered machines,the Xbox 360 and soon Zune.
For everything else it doesn´t look that great and even on older Windows versions one has to bundle the Compact .Net framework and other things with the game so the user can run the game.
Since those things buckle up to a filesize that is bigger than most entire flash games and many smaller casual download games i wouldn´t suggest making standalone download pc games with it,unless its about retail games or you know that its a game type for an audience that is fine with loading bigger files.
Next up,yup,also no web deploy as of right now with this.

Let´s go through some more strengths and weaknesses quickly:

plus:
-Performance:
In comparison to flash way better runtime performance. Not just code is executed much faster, on the artside one has access to pretty much full directX 9.0c capabilities and using the graphic card nicely that way, so graphic performance and feature set is far beyond flash limits.
This also includes shaders etc, the nice mubmo jumbo that 3D cards are known for.
-Testing and deploying to 360:
One can get an optional Creator´s club membership (99 bucks a year) and then connect the pc and a usual retail 360 console and then deploy a game built to the console and test it there right away.
Anyone who has played with the idea of getting into console biz and knows how troublesome it can be just to get a devkit (besides the costs also coming with that) will know how big a leap this is. Its really something quite special.

Some downsides:
-No visual editing of assets. Yup, don´t expect doing timeline animations,vector graphics,writing text or anything else related to graphical work here, you develop in a code editor with something like a library list added where one can see the names of assets, don´t expect anything more.
-Its still a young thing, there aren´t importers and processors for various filetypes, mesh importers have various limitations and there´s not even support for some things taken as granted and basic with flash, like handling vector graphics or even just writing text.
In case of text one has to write it using a bitmapfontplotter,similar to how one can do it with bitmapdata in flash if one wants to.
One can write one´s own stuff and use things others made available and the community constantly does release nice stuff, but yeah,don´t expect nice functionality for everything out of the box.
-Selling games: There are various limitations when its about wanting to sell a game. See above regarding making download or standalone pc games. Next up on the 360 you can test your game but then there is no way to let the general public play your game on the 360 yet. You can pass it on to other creator´s club members but not to the general usual 360 owners yet.
(MS plan seems to be to do several things there:
1.: Introduce a portal system towards the end of the year which will allow usual xbox 360 owners to download and play xna games for free on their system
2.: some higher quality games will surely get the special treatment and get published as commercial titles on XBLA then
3.: One will probably be able to buy xna made games for zune on the zune store, i can´t personally test this as zune isn´t available here.
)

Oneself could of course also do a very nice demo XNA game and then contact any Windows/Xbox 360 publisher with it, chances aren´t that bad they might be interested seeing the seamless deploy on the target platforms when its about retail games or XBLA games released by a publisher.

So seeing the restrictions i listed above to me personally XNA is very teasing regarding making games for the 360 which can then be tested and in most cases run right away on the 360 (sometimes small changes have to be done between Windows and 360 versions), it has the potential for getting into console biz easier and in case one is into getting in there or even later learning C++ the language is closer, the reputation better among bigger companies and publishers.
On the other side when its about not just learning things for fun or future usage or getting a retail pc or (download) console game gig going but instead making money for sure with your game right away once its done then things look worse than with flash right now. With flash we´re meanwhile used to the many portals and various other ways to make money with a game, if you´ve got a good game chances are high you´ll make some money with it in one way or the other if you know what youre doing. With XNA chances are there but not as secure and not that many and not with pretty much instant result.
Next up regarding the development tool and asset handling limitations as of right now one could do a 2D game quite similar to a bitmapdata based game setup in flash, when its about developing 3D games the raw power is there but tools for handling and editing it are not really that evolved yet,so expect lots of custom own work just for handling the stuff before the step of writing your actual game or during it.