You need to engage the users. You need to watch them and note down their assumptions, blunders and sources of frustration. You need to go around them not by trying to educate the users but by trying to educate yourself. You need to forget your own preconceptions.

Which is why most Linux window managers suck. The only ones that do not are those that mimic something that has been developled by a big company with a lot of specialists whose only job it to study users' behaviour, use patterns and workflow. Something like Windows 95/XP/7. That kind of layout - mimicked in KDE, Gnome 2 or LXDE, to name but a few - is not just a random concoction, put together because Microsoft could not come up with anything better. It is a result of a careful study and is the ideal solution for a desktop that has survived mostly unaltered for 15 years.Just look at the disasterous results when some developers or PR executives decided to move away from that. Unity, Gnome 3 and Metro (in that order) are glaring examples of how not to do a desktop environment. Their only justification for existence is that they are different, but literally no thought has been put into their design other than the looks.

Sadly, the same layout would not be as suitable for a tablet. Fat fingers need a different approach than a precise mouse cursor or a stylus. Luckily, there are other desktop solutions that are better suitable for a touchscreen interface. EasyPeasy comes to mind, for example. Or Gnome 3. As bad as it is on a desktop, it may actually be not too bad on a tablet.

However, the more I think about the topic the less I am convinced that the idea itself is sound. I think we are already zoomed too close in. We need to take a big step back and ask, what do people need a tablet for? Web browsing, watching Youtube videos, reading emails, playing games? There are already existing solutions that may not be perfect but are Good Enough™ and are already established too well to try to push them out. Some serious field work? But what? What can an alternative tablet offer that cannot be solved with an existing solution? Remember to replace a well established solution you cannot be 50% as good (that was Jolla's problem). You cannot be 90% or even 110% as good as the existing solution. No, you must be at least 3 times as good.

Other than the server niche, can you name a single thing where Linux offers the end user a three times better solution than Windows, Mac, Android or iOS? A three times better browser, email client, office suite, video editor...? Linux is mostly used by geeks and yet it does not even have a good programming IDE, for Pete's sake!

I am afraid that to make a successful Linux tablet, you first need to make a successful Linux desktop. As we all know, that has not happened in 24 years. It looks like we still have a long way to go