My personal rule of thumb - as few redirect jumps as possible. Three main reasons:

1. User journey + Browsers - Sometimes when there are too many redirects taking place, some browsers find it difficult to follow through and would simply not load the page.  Also, even if there were only 2-3, the browser may load, but users on slower connections may find it tiresome waiting for content to load.

2. As ThompsonPaul highlights, you COULD lose some link value due to dilution through 301 redirects.

3. Multiple 301 redirects are often used by spammers and I foresee in the near future these causing a lot of ranking headaches. The older the site, the longer the chain might end up - for example, imagine you had a product at:

https://domain.com/product1
Links to that page exist at domain.com/product1

The journey would be: domain.com/product1 >http://domain.com/product1 > https://domain.com/product1

Now imagine a year down the line, product 1 is discontinued and you decide to redirect https://domain.com/product1 to domain.com/product2

Imagine your journey now:

domain.com/product1 >http://domain.com/product1 > https://domain.com/product1 > domain.com/product2 >http://domain.com/product2 > https://domain.com/product2

This could carry on indefinitely in the lifetime of the site...

Best solution: Decide what version of the site you want to use and simply try and use only one redirect, not a chain. Periodically check for chained redirects and resolve as you go along. (I try and do this bi annually).

If you switch the page then there is a slight risk, although you are doing it for the right reason. One of the better and safer ways of doing this is to serve an image or a small block of text that directs users to the better matched section, and only geo serving that content on the page, not the whole page.

Personally I prefer the .com versions as they van be future proofed. It's hard to rank a dot travel domain.  Dot org rank ok, but I think loose their branding value. If you are building a long term pure white hat strategy, I would focus on a decent dot com. And register the front org and dot net versions too.

If you are looking specifically for link analysis tools then a pretty good alternative is http://linkrisk.com/

I have managed to get many penalties overturned based solely on using them as an analysis tool.

1. I wouldnt advise redirecting the robots.txt to redirect to home page. It seems that they hve a dynamic 404 redirect system - which when a URL doesnt exist the site redirects it to home. There are god and bad points about this strategy, hoever I would prefer NOT to do it.

2. Re getting site indexed - no it wouldnt hurt them, but would give you much less control over the robots directive, in case you want to add custom instructions. If Google crawlers cant get to it (as in its not user agent cloaked to allow the google bot) you will not be able to do so (eg excluding pages from being indexed via robots wont be ossible).

Totally depends on the niche I am afraid. however for small businesses, I always advise takingthe manual route - although old, the strategies here still work:

http://www.seomoz.org/blog/small-business-link-building-part-a-analysing-opportunities

http://www.seomoz.org/ugc/small-business-link-building-part-b-grabbing-the-bull-by-the-horns

To start with if the forum ISNT indexed, get the URLS changed, and try and get all the titles customosed so they arent generic across every thread.

I suggest a subdomain for forums, as you point out, it is seen as a separate site. you can always flow strenght from the forum into the main TLD by cross linking, which I think is much better to do, however some may disagree with this.

If I were in your position, I would prefer to release section by section - a massive blast of content would be more of a flag then a slow release - this is my personal opinion.

Re threads - I would follow/noindex all threads tht arent on the first page of the post -  that way, the juice follows and gets crawled. Adding relcanon to those may also help preserve the strength...