I have this observation of mine, that good (usable) user interfaces on the web are usually at their best upon the launch of the web site and some time after that, and then, as soon as the business behind the web site becomes successful (if it does become such on the first place), the great UI gets gradually ruined over the course of the success of the business (especially if that is an online business amassing a sufficient number of users to make it either a monopolistic or oligopolistic player). Do you have any observations confirming or refuting mine?
Disagree. I think Amazon, for example, has done a pretty good job in terms of improving the information architecture (IA), search and personalization of their interface over the years. I suppose this is the kind of "monopolistic" company you're referring to? It is true that IA often gets more bloated as companies add more elements to their nav over time. But this can be kept under control with careful organization and user research techniques such as card sorting.
I see little to support your theory, and plenty to contradict it.
Interfaces tend to get more complex over time as a company grows, but that is generally because the functionality that interface needs to support also grows more complex over time. Small companies do one thing, so their website can be simple. Large companies do many things.
Complexity is orthogonal to the quality of the interface -- if anything, more complex interfaces give more opportunities for good UX design to happen. There's a tremendous amount of functionality squeezed into Amazon's homepage, for example, which would have been completely overwhelming in their early UX design.
For many of the truly large-scale websites it's not really possible to make a fair comparison between their original design and their current one; the truly godawful appearance of (for example) Youtube, Facebook, Google, and Amazon at launch can mostly be blamed on the limited capabilities of the web at the time -- and, honestly, the comparatively limited skills of the web design community in the olden days. (I should know; I was one of them.)
But it can be instructive to go through archive.org and watch websites designs evolve over time; you can really see how the rough edges get honed off and incremental improvements get made. Just to take a couple of (very different) examples -- I'm not going to try to critique the entirety of their designs, just focus on one aspect in each case:
Amazon circa 2006 had a triply-redundant "browse" sidebar, "See all 34 product categories" tab (itself a compromise after years of gradually-increasing-in-number individual tabs), and product category pulldown menu. A few years later they'd got rid of the tabs; and a few years after that tucked the sidebar out of the way. By 2014 or so the category pulldown and search field had merged into a tidy unified control.
Stack Overflow has had a relatively stable functionality set throughout its lifespan; its design evolution has been more about incremental improvements than about major redesigns. It launched with comically amateurish typography. It took several years for them to start figuring out that not everything needed emphasis; over the next few years the typography gradually evolved towards readability.
I could go on, but you get the idea. It's a worthwhile exercise to go through the history of pretty much any site you may think of; sometimes there are missteps and unfortunate trends, but generally it's difficult to find an example of a major site buried in its own cruft (imho that's more the territory of small-to-midsized businesses that let the developers or marketing teams do the UX design by accident, or of companies that allow monetization or major changes of business plan to take over the UI.)
(I should clarify that I'm in no way saying this is universal -- designs don't always improve over time; sometimes companies do make missteps, outsource to the wrong design agencies, fall for fads or trends, etc. I'd say it's generally true that interface quality tends to improve over the long term.)
Classic Small Business vs. Big Business Attitude
I have had a theory for many years about this. It is NOT specific to "Web Interface." In fact, the typical small business will have a minimal web presence and therefore may not have much of a "Web Interface" to speak of at all. Now my theory:
All businesses starts small. Most either stay small or fail. A very small percentage become big businesses. For an example, compare how many small grocery stores, pharmacies, etc. there are (though admittedly not as many as there used to be as many have been swallowed up by bigger companies or closed due to competition from bigger businesses) compared to large grocery store chains or pharamacy chains, each of which can include hundreds, even thousands of stores. So there is actually a relatively small group of big businesses to compare.
In my personal experience (yes, opinion based), there are companies within each industry that have managed to take the "ideal" small business attitudes regarding customer service (personalized, "customer is (almost) always right", easy returns), employee treatment (not so much salaries as benefits, flexible time off, high morale), community involvement (e.g., supporting charities) and other factors. A high-quality web interface is simply the latest of those "customer oriented" factors.
There are other companies that have managed to grow large, either based on those same attributes but "losing" them along the way at some level, or purely on price competition (where, depending on the market and other factors, people will buy despite poor customer service), government sanctioned near-monopoly (whether by government regulations, shaky patents or other methods) or other methods to become REALLY BIG without those cherished small business attitudes. At a certain level, a business can become "too big to fail" and survive for a relatively long time despite these problems, including (in the modern era) a web interface that is far below expectations. (Though for anyone who thinks a business is truly "too big to fail" can simply look at the downfall of Sears after over 100 years as a leader in US retail.)
In fact, I have actually seen some big businesses with fairly decent web interface, but unable to compete successfully against other online retailers due to other factors, and with all of the other problems mentioned above conspiring against them, in the long-term, in the bricks & mortar world.
While I won't list my personal opinions as far as which big companies fall into each of these categories, as that is arbitrary and subjective (except of course that my answers are the right ones!), I think most people who do a lot of shopping at multiple large stores of a particular type (department stores, home improvement, grocery, computers/electronics, pharmacy), can come up with a categorization of which ones have a "small business attitude" based on multiple factors, including web interface.
The end result is that "quality web interface" is only one of many factors and does NOT necessarily change (for better or for worse) with the growth of a business.