Lightroom is much trickier as the library itself is the result. Photos have ratings, non-destructive edits, there are collections, tags, etc. Without ability to use the software the "result" is not accessible and effectively taken away.
This statement is just plain wrong. Over and out.
Very cute - but how about a discussion rather than just a put-down? I know little about Lightroom, beyond the fact it's awkward and clunky to use (compared to more recent and more intuitive software), so I, for one, would like to know how and why it's wrong.
Sigh. Not wanting to add to the sarcasm, but before throwing shade on Lr, or any parametric image editor and DAM, one should know how they work.
So, in a nutshell, here goes.
First, take the metadata the ratings, tags (more properly keywords) and similar stuff like titles, instructions, locations info, copyright, model name, and so on. Most of that is standardized IPTC metadata that has been around AFAIK since before Lr. And it's completely interchangeable across computer platforms. Lr (and other DAMs) can store it in its database, or in the image file or its sidecar. Or both. Pretty much everything reads it, even Spotlight on the Mac. So you could lose Lr, drop dead, and your successors could fire up your computer, grab an image, and note that you'd titled it "gravestone suggestion" or something.
Some things aren't standardized in metadata. Like maker notes in exif. That sometimes includes ratings. Labels. Not all DAMs can do hierarchical keywords, but several have adopted Adobe's method (Graphic Converter, Mylio I think, Photo Mechanic, maybe Digikam, Xnview MP, iMatch I think, and probably others. Pick flags are often proprietary.
And while most all DAMs have albums, collections, or whatever they choose to call virtual folders, these are not interchangeable. But note that usually every aspect of a folder, album or collection hierarchy can be duplicated with either hierarchical keywords, or even flat keywords (tougher, since it requres a lot of AND-OR-NOT foo for searching).
Other unique to the program stuff includes editing history (maybe using layers in Ps is a better choice), virtual copes (maybe a real copy is necessary), and really unique stuff like publishing history in Lr, or say sessions in Capture One, and so on. But this stuff generally isn't as much a problem as the basic structure you've set up, and again, an awful lot of that can be essentially written into the files (a more robust solution anyway).
That then leaves the parameters themselves, the zeroes and ones of adjustments to the image content itself. Some of these can carry to other non Adobe programs from Lr, to say Mylio and Digikam. But it's rarer. And really uncommon for other PIE programs. So it's a problem for all of them. One can choose to export a "print" of the adjusted image, like a TIFF. Or even a JPEG; some of us have found that re-doing the raw processing and adjustments in newer software often is about as fast using a JPEG as a visual template. Sometimes it produces better results. But it can be a pain, as when say a wedding client requests prints from hundreds of already approved images.
TL;DR: Any time you switch from a program to another one there will probably be pain. It can be minimized at least for organization and finding things, but to preserve a look of the content itself one needs to generate lots of TIFFs or other output.