Dating Jesus’ birth is partly tied to the dating of Jesus’ crucifixion (33 AD) and baptism (28 or 29). This is as it ought to be. Settling on the Lord's birthdate should NOT, however, be restricted by over-committment to worn out apologetic refrains aimed purely at defending a [reading of a] verse than have little or no practical concern for reconstructing an actual, historical event sequence.
Yes, dating Jesus' baptism is tied to stretchable interpretations of the word "about" and stretchable interpretations on the "fifteenth" year of Tiberius. Not too stretchable, thankfully, so we can at least settle on 28 OR 29. Following Cheney, I conclude an induced passover in Matthew 17 (tax season) gives us 4.5 years between the major events, so the baptism falls in (autumn) 28 AD. Unfortunately, and for various reasons, a lot of traditional christian scholarship has tied itself to a baptism in 29. That causes enough problems with event-sequencing on the Lord's ministry years, but what's worse is that it leaves a historical nativity completely out in the cold.
On the feeling that we shouldn't stretch "about thirty" to beyond 34 years, traditional Christian scholarship has (de facto) excluded 7 BC (and earlier years) from consideration. If there's another reason, I honestly cannot detect it. Unfortunately, the years 6 and 5 BC give us no plausible justification for reconstructing a Roman census in Herod's Kingdom - 6 BC saw a turnover of Syria from Saturninus to Varus and 5 BC gives us nothing at all to work from. (There is more to say on these points for another time.) On the other hand, a baptism in 28 AD allows a 34 year old Jesus to have been born in 7 BC. (There was no year zero.) And in case you didn't know, 7 BC has quite a lot going for it.
The case for 7 BC has been made on this blog, and will be made again. But the point just now is that traditional scholarship has settled for assigning years to something that has never been fleshed out in full historical context. Yes, at some point we all leave some questions to faith, but a whole census? Responsible faith-based historiography must at least try to reconstruct some explanation for why Augustus Caesar would have mobilized hundreds or thousands of Roman soldiers into Herod’s sovereign Kingdom, especially since it was merely to count heads (and not property also, unlike the census of 6 AD).
Failing or refusing to attempt even modest reconstruction gives tacit support to the notion that events in scripture may be unhistorical. Believing they may be historical should rightly require us to support that belief with historical evidence, if possible. Some defenders of our faith have unfortunately been straining out gnats by defending interpretations of verses, when the entire historical content of Luke 2 has been mainly ignored. To me, that camel has not been swallowed, it has been pretty much left for dead.