Initial thoughts for counterarguments:

1. Not comparable: Canonical, as a company only had one shot (or thereabouts). KDE, as a self.sustaining , volunteer-driven non-profit can take many shots. If we don't succeed now, we can try further on.
2. Kirigami will allow developers create apps easily for all platforms, including Plasma Mobile. We are already seeing many KDE apps ported over to Plasma Mobile. Plus things like Anbox may help us run Android apps on Plasma Mobile.
3. I don't think we can counteract the argument, but we can explain why the price is what it is: expensive components for a custom phone, with a small production run makes it impossible to make it cheaper. HAVE SOME FAITH!!!! (Plus innovation and privacy come at a cost)
4. Linux broke the Microsoft monopoly, Android broke the iOS monopoly (don't remember when Android was the underdog? Well it was.), Samsung broke Nokia, etc., etc.. Thinking that markets are static is short-sighted.

One point that I think is fundamental to make is that we don't want to be as excluding as Ubuntu and Sailfish did (note that we have several apps ported to Android but none to either of those, because they're far more restrictive!!!!). In principle any Linux application should work on the Plasma Phone so rather than trying to push down a very specific technology we can leverage the whole Linux platform as whole.

Simple. Any packaging format (flatpak, snap, even appimage should work if Purism doesn't have restrictions to add), any toolkit (Qt, Qt+Kirigami, GTK, enlightment, electron, raw OpenGL or vulkan). Instead of making a Linux platform for phones, we are bringing the GNU/Linux we know on a phone. This could make a big difference. And also if anbox one day happens it will be welcome, because if it works on your linux, it works on your Librem 5.

Then it boils down to "will I have whatsapp on day 1" which sucks.

Good points, @apol. Will take them into account.

One thought that comes up with me is that we don't necessarily target the main public. For the project to succeed at first, we don't need general audience size sales, we need to find one niche that we can serve really well. Thoughts:

We basically need one or at least a few organisations that want to communicate privately, without prying eyes from the U.S. or anybody else. Here, the small footprint of apps is a plus (employees will be less tempted to use the device for their private communication, whatsapp with their friends, facebook, whathaveyou).
Governments could almost be forced to use something like this since they have an obligation to deal with data privately.

If we could just get one or two organisations to buy in, we're basically there for the first line of products.

This strategy also counteracts all the critique you brought up.

Yes, a niche market like the one you describe would be ideal for this project. Maybe we should start brainstorming organisations wealthy and patient enough to find this project attractive.