My expertise is in Oracle - readers don't block writers and writers don't block readers. Infact writers don't block other writers unless they are updating the same row. I think MySQL works the same way as Oracle in this regard.
I find it hard to imagine building a performing web application on top on a database in which writers block readers!
I don't know.
One thing I do know is in the places I've seen SQLServer used this wasn't a problem. In StackOverflow #17~ http://itc.conversationsnetwork.org/shows/detail3792.html Atwood mentions using LINQ a lot, instead of raw SQL or code they have written. I'm wary of generated code from MS tools. I wonder if this might be the problem. You can read the transcript here. The podcast is pretty good. Spolsky the Morcombe to Atwoods, Wise ~ http://en.wikipedia.org/wiki/Morecambe_and_Wise
"... I'm guessing that MySQL, which grew up on web apps, is much less pessimistic out of the box than SQL Server ..."
That one sums it all up. But does it summarise the problem? SQLServer works well enough out of the box.
So if he's just inserted record #10, and is now trying to select it while simultaneously inserting record #11, there can be some problems with timing. It's pretty rare, but if he's running a program to blast a bunch of stuff in at once, I've seen it happen.
The other possibility is that he's inserting or updating foreign key requirements at the same time he's trying to select out that data; and he has the option turned on to enforce foreign key constraints. SQL will lock the dependent data while any updates or inserts occur to ensure the FKs are valid through the lifetime of the transaction execution on the foreign table.