Add CalcResult<T> class

[Vovanda]

• Provides the ability to use standard arithmetic operators
• Replaces the usual approach to throw an exception or return the null in case of an execution failure

[Konard]

I suggest you to look into the Integer<T> struct implementation that was removed from this repository. May be we can revive such struct by overloading its operators that will use our dynamically compiled delegates for basic operations. May be we can also rename it to Number<T> (1 letter shorter), but this will also mean we will have to implement support for floating-point numbers (float, double, decimal).

[Konard]

@Vovanda after this will be finished you can also try to compare our solution with existing one (https://github.com/linksplatform/Numbers/issues/31), or may be we can use some ideas from Jon Skeet`s implementation.