Technologies that had to mature before the tablet computers became practical:
Wifi networking.
Useful, I'll agree. Not "necessary". I ran cables throughout the old house in the mid-1990s, and was getting a stable 100MBPS connection from any computer from about 1996 until we left in 2012. If I had a guest and I wanted to provide them with WiFi, I'd turn on the laptops WiFi card and the last time I did it, they could get half the connection speed that I had through the cable. I was considering running 1000-base, but would probably have left it until the previous cable was 20 years old before replacing it.
Capacitive Touchscreens -- Most early designs used a stylus, which sucks, and had poor resolution to boot
I used a touch screen with a stylus. You might think that they suck, but I'm perfectly happy with them. My Psion used one (and I never lost one!) and my last - or last-but-one - phone also had a stylus (which I also didn't lose, until I lost the phone itself). You might think that they suck, but that's a subjective opinion, not an objective fact.
Low power but still acceptably fast processors -- A huge sticking point, lots of early tablets had extremely poor battery life on top of being slow
Yeah ... in 2000 I objected so badly to replacing the AA cells (rechargeable or primary) EVERY DAMNED MONTH. It was such a pain in the arse having to go to any shop in the world and buy two batteries EVERY DAMNED MONTH. It's so much better having to carry a charger (and the panoply of adaptors for the 5 different sockets that I meet most months) with me and having to recharge the device several times a day.
A touch enabled OS -- WinCE is terrible to use with a finger, and really pretty bad with a stylus. Symbian was never great. PalmOS was too narrowly focused on Palm pilots
When I discovered Symbian, I never felt the need to try a WinCE machine or a Palm machine. I just got on with using the applications and barely noticed the OS. Which is how it should be.
Battery capacity -- Battery technology has come a long way inetwork speed n the past 15 years. Early attempts would use NiCad batteries, which just aren't good enough, especially with the relatively high energy consumption figures from the old chips
See above comments about the horrors of a monthly battery change (Either NiCads, NiMHs or primaries).
The technology to make effective "tablet" devices was available in the late 1990s - Psion did it. To this day, it's a mystery to the community of "Psioneers" why they stopped manufacturing them, or why they didn't sell the hardware division as a going concern when they restructured to become a software-only company. If they'd continued ... well, the world is full of "if onlys".