>but they must be at least partially technical.

Having sat in many a meeting, partially yes, but these things are massively expensive. There is an equation, How much would it cost us to write a replacement that covers what we need versus how much does it cost us to use what exists that isn't efficient.

And this is where you miss the biggest part of the problem. It's the end users that bear the biggest part of the costs. Yes, there is an internal cost for their own developers, but that is comparatively small to the costs of their paychecks.

The next comes to management of the lightweight solution over time. If it's owned by a company at the end of the day companies are rarely interested in lightweight, they are interested in making the most money and quite often that means adding more and more features to accomplish lock-in.

Open source is more likely to keep a project remaining light, but to do that it's quite often by not accepting bulky features that would make companies more money. So you see where the catch-22 situation starts to arise from.