Where the focus on development will be entirely web-based solutions i.e.: (eCommerce/organisation management software) developed by a small team of about max 5 - 10 techies.
Taking into account;
Please suggest your solutions in this format (OS, Server, DB, Language and Methodology) i.e. (LAMP and XP) or (Linux, Apache, MySql, ROR and TDD).
Open Source is the way to go when cost is an issue.
Also, feel free to erase the word "waterfall" from your software development vocabulary. It simply does not exist ;)
There's no reason to avoid open source solutions if cost is an issue; these days, they're every bit as good as the best commercial products for any small installation (i.e., anything you'd spend less than hundreds of thousands of dollars on).
Myself, I'd go with FreeBSD, Apache, one of the good scripting languages (Python, Ruby, or Perl, in order of my own personal preference), and PostgreSQL.
On servers, FreeBSD seems a bit more stable than Linux, at a cost of having a slightly different feature set and being a tiny bit less pleasant to work with. Either one is fine though; Linux makes a better workstation, so there's choices all around.
Apache is a good server choice because there's nothing it can't do. You will never need another server if you go with Apache; even if it isn't the perfect fit (some other product might be a little faster or more natural for your current work), you know Apache will handle it and handle it well.
Any of the scripting languages are a good choice; PHP is okay, but Python/Ruby/Perl are such good alternatives these days that it's hard to see why anyone would choose PHP for a completely new project.
PostgreSQL is superior to MySQL in just about every respect (far better feature set, more stable, at least as fast). MySQL does, however, have the edge on replication. PostgreSQL can do it, but it won't be natively supported in the core until the 8.5 release at the earliest (expected in 2010). As well, if you expect to have complex queries, PostgreSQL is derived from Ingres, which was famous back in the day for its extremely powerful query parser, making it a natural fit for messy work.
Going FLOSS makes the most sense unless your programmers are already proficient in a proprietary technology. Why worry about licensing restrictions and support contracts and all that when there are proven technologies available for free -- and that arguably run the majority of the Internet.
That said, there are so many permutations of answers to the rest of your question... I think you'll need to provide a lot more information.
There are several advantages to Open Source here.
First, of course, is the price.
Second is the lack of licensing hassles, which may be more complicated in a developing country.
Third is the likelihood of additional security.
Fourth is that it will likely run on less sophisticated equipment.
Of course, if a closed-source app meets your needs right off, it might be better to buy it and take advantage of time-to-market and potential lesser development costs. (I have no idea exactly where you are, your finances, the cost of skilled techies where you are, or anything else like that.)
I'd suggest evaluating what you need, seeing what's available, and favoring open source if there's no compelling reason otherwise.
For your last question, without knowing considerably more about what you're doing and what your constraints are, nobody's going to be able to make a good recommendation.