As we all learned in kindergarten, every snowflake is unique. Now, studies have shown that is not entirely true (see here if you are marginally interested in the science of snowflakes), but close enough.
Uniqueness may be good for snowflakes, but lousy for financial transactions. At least part of the problem in harnessing private capital to fund energy efficiency projects is the lack of standardization across projects. Major financial institutions need big pipelines of medium- to large- deals to make the sector worthwhile. If each transaction is significantly different, it means that underwriting standards, loan documents, due diligence and so forth have to be created from scratch for each transaction.
The solution to the snowflake problem is to standardize the transactions. It is very tricky with energy efficiency, because each building has different features--age, system types, use, occupants, etc. Therefore, the technology needed to create the efficiencies also differ.
To break through the logjam, someone needs to develop one or more reliable algorithms that can be used to evaluate projects quickly, while still respecting the unique features of different projects. Transactions could be structured around the standardized model, and then tweaked for the particulars of the transaction. Without a reliable and replicable structure for financing energy efficiency deals, financing will always be limited and expensive.