Replacing my MacBook Air M1 with a ThinkPad T480

This is such a weird thing to say.

I have a 256GB (with easy/cheap room for 768gb) 28c haswell-e server sitting in my basement though too. Costs like 1/3 of a macbook pro if it comes down to it.

Rule #1: do a ton of research and comparison shopping to be certain you're making a good PC buy. These huge brands like Lenovo, Dell, Acer, etc churn out so many laptops, and the build quality varies wildly between them. For my process, I completely factored out most brands: absolutely no shopping for Acer, MSI, Dell, Toshiba, etc. They're just too much of a roulette, unless you know a specific model that fits your needs (like the Thinkpad!). At the time I was shopping I was only looking at Asus, Alienware and Gigabyte (not sure if they're all still good brands in 2023).

Rule #2: be able to do basic small maintenance. I ended up buying a Gigabyte laptop that's still running perfectly 5 years later, and it IS a gaming laptop with all the heat / cooling issues that implies. I've had to replace the SSD, the battery (heat degradation), didn't replace but had to oil and reseat the fans, repasted CPU / GPU, and make various software tweaks. A computer shop could do the same, just like the Genius bar, but it just makes things so much less stressful when you can do the kind of quick maintenance that Apple so strongly discourages.

Upgradability is not free. It has a weight and volume cost. Depending on implementation it can also have a performance cost - for example a user-swappable battery means a 30-40% reduction in battery capacity OR a corresponding increase in weight+volume. For another example SODIMM slots take up space and are larger than surface mount soldered DRAM chips... I mean that should be obvious right?

Conspiracy theorists are gonna see conspiracy theories.

Engineers look at ancient "luggables", imagine a future where Star Trek pads exist, then incrementally iterate reducing power consumption on 200 different individual components, shrinking the volume of 27 different assemblies, and making thousands of difficult choices - all of which has to yield a sellable product today so there will even be a chance to improve in the future. There is no magic. You can't wave your wand and make the future week-long battery life laptop that weighs nothing appear out of nowhere. Somewhere along the way you have to decide if you are going to keep the space for the extra connectors and SSD controller or if you're going to integrate everything to shave off some weight and reduce volume, along with cut power consumption by removing a single chip. The nay-sayers claim the gains are minimal for those specific changes but you can make that claim for most of the changes made along the way to get from a luggable to the modern MacBook Air. Some people really needed SCSI ports and are still angry they're gone. If everyone listened to those people laptops would have the weight of bricks and get 20 minutes on battery.

History has proven those with the engineer mindset are more likely to be successful than those with the conspiracy mindset but it's up to each person to choose.

(Of course you can also make a living filling a niche - the world is large)