Reworked the sample I saw floating around to be closer to how I'd write it. Did not fix bugs.

EnglishMotherF.cs

// Two tips from Brian Kernighan:
// 1.  Say what you mean, simply and directly.
// 14. Use symbolic constants for magic numbers. (Applies to chars and strings too.)
//
namespace System
{
    public static class SentenceParser
    {
        protected char BLANK = ' ';

        public static string TrimSentence(this string @input_string, int length)
        {
            if (BadInput(length, @input_string)) return string.Empty;
            if (length >= @input_string.Length) return @input_string;
            if (NextCharacter(@input_string, length) == BLANK) return NormallyParsedString(length, @input_string);

            return SentenceEndingAfterPreviousWord(@input_string, length);
        }

        private static string NormallyParsedString(int length, string @string)
        {
            return @string.Substring(0, length);
        }

        private static bool BadInput(int length, string @string)
        {
            return length <= 0 || string.IsNullOrEmpty(@string);
        }

        public static string TrimSentence(this string @input_string, int length, bool useEllipsis)
        {
            return TrimSentence(@input_string, length) + OptionalEllipsis(useEllipsis, @input_string, length);
        }

        private static bool OptionalEllipsis(bool addEllipsis, string @string, int length)
        {
            return (addEllipsis && @string.Length > length && length > 0 ) ? "..." : string.Empty;
        }

        private static string SentenceEndingAfterPreviousWord(string @string, int length)
        {
            var lastSpaceIndex = @string.Substring(0, length).LastIndexOf(" ");
            return @string.Substring(0, lastSpaceIndex);
        }

        private static char NextCharacter(string @string, int length)
        {
            return @string[length + 1];
        }
    }
}

Pondering the names for NormallyParsedString and SentenceEndingAfterPreviousWord cost me the most mental burden here, so I decided not to rename them. I removed articles (the, an, etc.) where possible to reduce noise.