One of the biggest examples of a constraint that I feel has benefited my organization & product is a requirement that everything must start out using SQLite. To this day, ~6 years after this one directive, we have not yet found any actual need to move off this approach. The amount of time we were able to spend focusing elsewhere has actually made us viable as a business.
There was a period where our organization was essentially unconstrained on tech (think microservices, tons of repos, mapping/API hell, etc). These were very fun times for the novice engineer. For everyone else it was a living nightmare to watch the weeks slip by without an ounce of additional business value being added to the product.
I really, really like this idea. After experimenting with SQLite recently I was quite impressed, and I can easily see how using it to abstract away the database layer for another day makes a world of sense.
Also at the language layer, constraints are a big win to me. I used to love Perl, don't get me wrong, but Erlang with its opinionated architecture (async, immutability, assertions everywhere) seems like a much saner way to develop a serious program.