Back then when this first was suggested and a program was made by ShunterAlhena, I already made such a 'draft' and the very bare boned basics for such an application.
I can assure you that such an application is far, and I mean way faaaaar, from being strait forward. And such a 'draft' isn't that easy to create either. There are extremely, and I mean extreeeeemly, many things you need to take into consideration.
To give an example out of the many: some controls do not have ampersands originally, but they can have them, eventhough they are not visible to the user in Plus! or eventhough they are not used. It isn't enough to just assume that only certain options will have them. You need to check every key in the whole translation. There are also even keys which are not listed in the main language file but which are still perfectly possible to use in translation. Some same ampersands can be used although the controls appear on the same page. etc...
Also to keep this up to date, you need to put a hell of a lot time into carefully revise every!!! option, link, button, control, etc in Plus! each time a new build is released. And I can assure you that doing something like that take a very serious long time.
An ampersand collision checking software is only usefull when it is absolutely 500% correct and very decently made. Otherwise it will give a false feeling. As events in the past have shown too frequently, people will rely on such applications too much hence it should work without the slithest bug or there shouldn't be one in the first place.
The time it takes to manually check ampersand collisions versus the time it would take to create such an application and updating it each time is very very very high. In fact, you would put more time in updating the 'draft' than the time needed to manually check a translation file. Hence why I also put this project in the freezer (well actually it has been put in C:\Program Files\MSN Messenger\Patches\MsgPlus\Translations\Tools\Validator\tests\
and maybe, just maybe, if there are enough translators complaining about the workload for checking this and too many collisions in other translations are found, I might consider unfreezing this
)
Not saying you should not do this, but I'd suggest you better put some time into creating something else.