Scientific correspondence has always been important – as part of the dialectical and hermeneutical process of interpreting data. The value of scientific letters recently dropped for several reasons

• Multiauthor papers may already have a balanced view
• Formalized discussion sections already contain most arguments
• There is no measurable “impact” as letters tend to be less cited

Having two accepted letters (in JACI and Gut) I even received comments like “braggadocio” or “blowhard” – nevertheless I would like to defend this scientific tradition.
Biomedical science is increasingly data driven with researchers neither aware of the historical nor current multidisciplinary context. I think that letters are not only fun to read for their brevity but may also carry important information if carefully selected by editors, yea, yea.