Performance and quality are rarely at odds with legitimate business interests or needs. In fact, if you sleight performance or quality it is exactly equal to sleighting future business needs or interests.

The trouble is when illegitimate (generally political) business concerns are allowed to be trumpeted as more important than legitimate business concerns, or when illegitimate views about how to discount the future value of quality are permitted to dominate the comparison with extreme short-term opportunities.

Performance and quality always take time and money. It's pretty hard to convince management to actively invest in something that doesn't add features to the sales pitch. The argument that long term it makes it easier to build more features somehow doesn't seem to work.

If you're doing anything on the web, Performance should be an easy sell to the business/management. There's a ton of solid data showing faster websites = more conversions which = more revenue. Then factor in hardware and licensing cost reductions by handling more capacity on less infrastructure, and you have another clear financial win. Honestly performance work is one of the easiest places to show a clear ROI.

"In fact, if you sleight performance or quality it is exactly equal to sleighting future business needs or interests."

This assumes that you are building something you will be using in that future which for most startups is not true. For most it will be a better trade off to iterate faster now and solidify the system once you know for a fact that it matches actual needs.

Most start-ups are deemed by the market to be not useful. I'm not sure it's informative to draw lessons of the value of emphasizing quality ahead of short-term, less substantive "results" from the population of start-ups.