I should probably leave the topic, because at this point I'm just rambling, but I can't help myself.
Your description of the job:
Are you sure that you know what I mean? It would simply require a resizable frame function that loads the chat script in it, and a menu that requires cookies. Not that hard, if I knew where to start.
My description of the job:
Hotlinking the thread url to the menu would require a little research into the forum system and the current template used. You would have to pull the URL of the thread from the page. Then you would have to create a UI or API or download one so that all the hotlinked thread urls would be editable, and then renameable, and then store every single one in a cookie, or perhaps several: You would have to save information regarding the name of the link, the URL of the link, and perhaps the position on the page. You would have to either make 2-3 cookies per link or a single cookie which you would parse with a regular expression to take all the needed parameters from. That step could take 2-3 hours. 2-3 hours = $200-$300. (Don't quote me on that, its a very rough estimate, I'm probably off).
The resizeable frame with the forums chat should be easy, but it too would require a little JS scripting. Best case scenario: The chat script can be loaded into the frame without conflict. Less than an hour - $50.
A news ticker with the latest threads. That is the killer, IMO. You would first have to research the forum system and check their database system. Then you would have to create a script that would query their database and check for the latest threads, and then display them in a ticker. <marquee> is considered to be bad web practice, but then again, so is frames, so that shouldn't matter. An hour or two: $100-$200.
Transparency on the menus, I believe, would be worthless with a frame; all the content of the main page is pushed over, not hidden underneath. In any case, it would require css as well as hacks for IE. Less than an hour: $50 at most.
On top of all that, not a whole ton of people are even familiar with greasemonkey, so either your chose of prospective coders will be limited or you the coder will have to learn to use greasemonkey to do the job, which will add time (and so cost as well)
Total time: 4-6
Total price: $400-$600
Again, rough estimates. You get what you pay for.