As in the Red Edition, the choice of string library (issue #1) has been the most controversial.  There was no majority vote cast in the Red Edition, so the issue is being reballoted.  Currently, the index-based SRFI 152, which is meant to be a simple basic string library, holds a majority position, but only by a single vote.  There is a strong minority for the original SRFI 13, which is a superset (with a few deviations) of 152.  SRFI 130, which is cursor-based, has only a single vote.  Three write-in votes were cast for SRFI 140, which I excluded from Tangerine because it provides adjustable-length strings.  These, like all other features that can't be implemented (at least minimally) on top of R7RS-small, have been postponed to the Green Edition.  I voted for SRFI 152.

Issue #4, supplementing the Red Edition's SRFI 127 generators with their dual, accumulators, is substantially beating the alternatives of status quo and no library.  Issue #6 is about bitwise operations on integers, and the comprehensive SRFI 151 is dominating the R6RS alternative.  The same thing is happening with fixnums (issue #7) and flonums (issue #8), where SRFIs 143 and 144, both supersets of R6RS, are getting more support than the R6RS alternatives. SRFI 160 is a superset of SRFI 4 that provides homogeneous vectors (issue #10), and it too is winning, though by a lesser margin.  Surprising to me is that for the combinator-based formatting library (issue #11), the combinator-based SRFI 159 is in a majority position over SRFI 48, the traditional template-based (as in Common Lisp) alternative.  Essentially all the remaining issues are yes/no/abstain, and yes is dominant all down the line, though a little less so for ratios (issue #13) and exact complex numbers (issue #16).  I voted with the majority for all of these except exact complex numbers.

So what is happening is that people are voting for more rather than less, as with the Red Edition.  This encourages me that I'm going in a sensible direction with the large language.