[acdha]

This is a valid concern but your comment really highlights a big problem with these discussions: your wording implies that this is universally bad but your examples are both things where it's impossible to say whether a given move is a win, loss, or wash without fairly detailed data. An increasing percentage of people are either mobile-only or mobile-primary, so making things mobile/tablet friendly is probably a good idea for most sites. Similarly, customization has significant training and support costs and removing infrequently used features to add things more people care about is a classic business trade-off.

[fourthark]

There's no real reason for a customer to care about these trade offs. They want to get work done, nor worry about the business model behind one of their tools.

[acdha]

That's looking at the problem backwards: if you have one customer who has a bunch of customization requests you need to weigh their business against what other work you could be doing with the same development time. It's not a win if you keep one customer but lose others to a competitor whose product is easier to use or cheaper to develop, especially since I've probably seen at least a 1:10 ratio for arcane features and customizations which a customer swears up and down are mission critical to things which actually are — usually it's more like one guy doesn't want to consider changing the way he works in the slightest until forced.