Just to pile on about how bad this is (and I know this was not your idea) starting with this sample data:
IF OBJECT_ID('dbo.table1','U') IS NOT NULL DROP TABLE dbo.table1;
IF OBJECT_ID('dbo.table2','U') IS NOT NULL DROP TABLE dbo.table2;

CREATE TABLE dbo.table1 (nameId INT IDENTITY, phoneNbr VARCHAR(20), email VARCHAR(100));
CREATE TABLE dbo.table2 (nameId INT IDENTITY, ContactInfo XML);

INSERT dbo.table1(phoneNbr,email) VALUES ('555-444-3333','JoeBlow@gmail.com');
INSERT dbo.table2(ContactInfo) VALUES
('<contactInfo>
<phone>555-444-333</phone>
<email>JoeBlow@gmail.com</email>
</contactInfo>');

Note these queries:
-- relational
SELECT t.nameId, t.phoneNbr, t.email
FROM dbo.table1 AS t
WHERE t.email IS NOT NULL

-- XML:
SELECT
t.nameId,
phoneNbr = t.ContactInfo.value('(contactInfo/phone/text())[1]', 'VARCHAR(20)'),
email  = t.ContactInfo.value('(contactInfo/email/text())[1]', 'VARCHAR(100)')
FROM dbo.table2 AS t
WHERE t.ContactInfo.value('(contactInfo/email/text())[1]', 'VARCHAR(100)') IS NOT NULL;
Before considering performance, just consider how much more complicated the XML-based solution is. Which is easier to understand? Which appears easier to troubleshoot? Which structure seems easier to maintain? 

Now look at the execution plans:

... and this is just one record. Imagine trying to join some other tables, adding a GROUP BY or other basic SQL. The queries will be miserably slow and impossible to tune because the execution plan becomes insanely verbose. Just more food for thought.