MS have played their hand very astutely and have managed to capture a large fraction of the enormous gap between cost-to-producer and value-to-user. Keeping things that way is worth $100bn market cap to them, and challenging them is worth relative peanuts to anyone else. Of course they're going to win every time.
Our best hope is that someone like Google pours money into a fully-compatible fully-featured desktop replacement out of sheer spite.
DOC/DOCX compatibility across different products is a tire fire, there is no solution on the horizon.
As for Excel and Word - those are 80/20 replaceable for LibreOffice or what-have-you (definitely for SOHO purposes: it will work, mostly), but it took over a decade of concerted effort to get there. Making a powerful word processor or a spreadsheet processor is orders of magnitude harder than it seems at the first glance.
Well, let's do some back of the envelope calculations. Let's say you can tap an incredibly cheap source of programmers, so that you can employ large numbers of them for 100k / year - this includes buildings, equipment, but also support staff and managers and what not. That would give you 10k man years of work. Star Office/Open Office/LibreOffice has existed for what, 20 years? So that's say 500 programmers. I'm not sure what the head counts on the various teams working on that have been, probably less than 500, but not an order of magnitude less. So considering that OO hasn't been able to replicate anywhere near the functionality of Office (I'm talking not just basic document processor here, but things like the VBA environment, and Outlook, and Access, and all the programmable components, etc.), I don't think you can say that 1bn is all that much.
Of course you could say oh but we don't need all the legacy stuff Star Office once did, and we'd just be a 100 coders with no managers working from our bedrooms eating pizzas, and we'd only build the 20% of things people would actually use, and we have much better tooling now so it'll be faster, etc. You'd still have to convince people to switch, in order to save what - $100 / year? 150? That's what, 1/2 day of work, if that, per user for a company? Sure a saving is a saving, but I don't see how you can claim that Office or Windows are 'expensive', in the overall scheme of things.
Office, in particular, brings in something like $25bn/year revenue with presumably very healthy margins. For a product that has hardly changed in 20 years. It's a good product, sure. But a clean-room reimplementation of the bits that people actually care about surely couldn't cost more than $1bn.
The power of monopoly capitalism.