This is both true and false: a fall in the price of a bond reflects opportunity costs. An example: you buy a brand-new MacBook Pro (bond) today for $1800. The next day, Apple cuts the price of the exact same model by 30% to $1260 (interest rates rise, bond prices falls). On the one hand, you have exactly the same computer you had on the previous day (same stream of interest payments), so you could say that you haven't lost anything. On the other hand, if you'd waited a day, you could have gotten exactly the same thing for $540 less, so you could say that you're out that amount. And if you fall on hard times and need to sell the computer, no one will be willing to pay you $1800 for it, since they can buy it from Apple for much less, so in that case the opportunity cost becomes an explicit cost.
I'm not sure what part of my comment you're disagreeing with; I agree that bond prices can fall significantly in a short period of time. I was mostly disagreeing with the statement that such a drop doesn't matter if you don't sell.
Whenever this retirement finance stuff comes up people mostly just extrapolate USA markets from 1982 forward, or maybe at best something like 1919 USA markets. I don't think history is so linear for this to be useful.
Excel on Mac works fine, and we do some fairly sophisticated numerical modelling work.
In terms of Office, the bigger problem has been legacy Access databases. We solve this by setting up some VMs with RDP access for the Mac users. But there's also a couple Mac utilities that let you export Access tables to CSV which can then be loaded up into something sane.
A surprising number of people also use Project, Visio and OneNote and there's not really a good portable solution for those cases. I've been pushing people to use yED for portable charting (instead of Visio), OneNote is now available for Office 365, and people are finding Excel a good enough Gantt chart tool for many purposes.
Office on Mac is quite good compared to Windows, just not all of the products people expect to use are available.
The bigger picture thing is that this is going to happen and enterprises need to figure it out. Non-cross-platform applications and processes need to be abandoned. This also happens in reverse with remote Linux servers being easier to work with on Macs than via Windows (putty and winSCP sort most of the issues out, but they aren't perfect).
In my experience people really don't care about the desktop experience too much, IMHO Windows is better than OS X for most of that anyways, but they care about "getting shit done" and that means apps, and these days web-based apps. The OS just isn't that big of a factor so long as the things people need to GSD is available. They're choosing hardware and form-factor over OS and they pick the one they think will work best for them.