It's often the case that management - and in particular, VC management - spend way too much time focused on irrelevant details. For instance, Thomas Gagne writes:

Too many potential investors spent too much time looking at the technology. The answer was always the same: lots of scalability, too little system documentation. The technical due diligence reports would often comment about the Smalltalk in the "risks" section something like, "Using Smalltalk may make it difficult to find programmers, but by using it they won't need many anyway."

Why is the worry about technology use irrelevant? Either the startup being funded has a good product or they don't. If they engage in a long effort to rewrite it in an "approved" language, they'll almost certainly miss whatever market window they are going after. The phrase Thomas used next is very telling:

Many of those VCs don't exist anymore.

I rather expect that a lot of them "managed" their startups into oblivion by asking them to do stupid things. To follow the story:

When we finally found some real investors I was surprised they didn't seem too interested in the technology. They wanted to know it worked. They wanted to know it balanced. They wanted to know the president had confidence in it and me. After they understood the business plan and what the company was preparing to do I really don't think it mattered much to them what the code was written in. They weren't buying into language marketing or hype, they were buying into a commercial finance revolution. The system could have been written in 16-bit Fortran-IV and they wouldn't have cared as long as it was supportable, scalable and correct.

Couldn't have said it better myself. This reminds me of something that came up many years ago, back before Java was around. I was teaching a Smalltalk class with Russ Pencin, who is one of the two best instructors I've ever met. We had a C programmer in the class who really, really wanted to know how strings were implemented at the VM level. Russ' answer was brilliant:

Why do you care?

Why indeed? If the system performs well, what difference does it make? The same goes for funding, really - training developers in a language they don't know already isn't that hard. If it was, C++, Java, and C# never would have succeeded. Sure, the syntax is similar - but they don't really work like C. If it was that hard to train developers, we would all still be using Cobol and Fortran, for that matter.

Thomas also quotes Paul Graham - and this is one of the smartest things I've ever read:

"In a big company, you can do what all the other big companies are doing. But a startup can't do what all the other startups do. I don't think a lot of people realize this, even in startups."

If all you strive for is mediocrity, then sure - hire cheap developers and use what everyone else is using. If you want the chance to do better, you have to be different.