One computer: once you decide you want to use your smartphone, for all but the most basic sheets there are problems with Excel mobile preview, even with Microsoft's official apps.
One location: Excel's lack of a universal time (showing different times in for users in different timezones) have led to gnarly workarounds. There was an attempt to correct for it in the XLSX format (cell type 'd') but it has its own problems.
One person: Excel has no inherent sync strategy. You are forced to farm out to email (and the inherent filename versioning nonsense) or use a system like SharePoint (which has its own problems you notice when dealing with people in different timezones across the world, like accidental file locking and data loss)
Any solution, SaaS or otherwise, will have flaws. It might solve Excel's weaknesses really well, but will likely lack its strengths.
Excel is still going strong in 2017 to a large degree because its strengths continue outweigh its weaknesses and that people have found reasonable ways, or add-on tools, to work around it's weaknesses.
It will be interesting to see in the coming decade if niche industry solutions or an even more incredible swiss army knife were to come around and dethrone the incumbent.
What I mean is that Excel is essentially a write-only language. It's very quick to prototype something for a semi-technical person, but auditing / code-reviewing an Excel sheet later is essentially impossible. The ubiquitous manual loop unrolling doesn't help readability either.
It also scaled fine to thousands of employees.
It's a pithy quote, you'll get a very different reaction from a finance or small to med-sized business professional.
Excel is the ultimate maker studio that actually lives up to the promise of "build your own mini-app". I'm not saying it can't be improved on. It's just that the core model of flexibility is so compelling, pushing everyone into a SaaS app is invariably going to take away substantial power from the user.