There are no end of improvements that can be made. Sure, an output buffer on the unbalanced output would be nice. But no, it isn't broken without one. It works to specification. As noted above, it will almost always be used with the level set to maximum anyway. There will be a few users that need to attenuate the output and may notice a loss of low frequencies in some circumstances.
This thing sells for $56 US on most sites I see. Assume a 2.5 multiplier over COGS, that means the manufacturer is making them for $22. In the USA. Everything is razor thin. You make money on these devices by saving the cost of a single resistor. Adding an entire dual opamp for what is a marginal gain in only some circumstances, and circumstances where nobody seems to be complaining in actual use, just isn't going to happen. You have to ask what other improvements could be made, improvements that are arguably just as good or even better to make. They are all in competition with one another, and overall none individually will make a huge difference over the basic performance. The next step up in performance is going to require a whole set of improvements to justify itself. Buffered output, better power supply, better line drive, better line receiver design. Each one is a useful but marginal gain. Together you get to a nicer product. But the BOM will rise maybe $10-15, assembly another few dollars. Price goes up, relative price in the market moves, sales may suffer, volumes drop, margins need to go up. You could easily be looking at a $100 US product. Better? Absolutely. But not the same product, and not the same market segment.
A new design, and have the entire thing made in China, you could probably have that better design for a similar retail price. But that isn't what this is.