Is the set of `##ALT` top-level types in IDs *really* restricted to the 5 provided?

[adamnovak]

Here the VCF 4.5 spec says there are 5 possible top-level types you "must" use for the ID of an ##ALT header line:

hts-specs/VCFv4.5.tex

Lines 259 to 266 in c101c79

	The first level type must be one of the following:
	\begin{itemize}
	\item DEL Region of lowered copy number relative to the reference, or a deletion breakpoint
	\item INS Insertion of novel sequence relative to the reference
	\item DUP Region of elevated copy number relative to the reference, or a tandem duplication breakpoint
	\item INV Inversion of reference sequence
	\item CNV Region of uniform copy number (may be deletion, duplication or copy number neutral)
	\end{itemize}

None of those is R or M. But then in the next section it has an example using those, for IUPAC codes:

hts-specs/VCFv4.5.tex

Lines 285 to 289 in c101c79

	Symbolic alleles can be used also to represent genuinely ambiguous data in VCF, for example:
	\begin{verbatim}
	##ALT=<ID=R,Description="IUPAC code R = A/G">
	##ALT=<ID=M,Description="IUPAC code M = A/C">
	\end{verbatim}

So are only those 5 possible top-level types plus IUPAC codes allowed?