First of all, the typical Evangelical Christian will think that the Gospel accounts of the crucifixion and death of Jesus are sufficient to prove that Jesus was crucified and died on the cross. But Johnson would not agree with this assumption, because he has a more skeptical view about the historical reliability of the Gospels. Johnson compares the Gospel accounts about Jesus with the accounts that we have of Socrates, and he finds the Gospels to be more questionable and problematic than the accounts we have of Socrates:.
The problems facing the seeker of the historical Jesus are even more severe [than the problems facing the seeker of the historical Socrates]. Although the biographies of Jesus…were composed within forty to sixty years of Jesus’ death, that is still greater than the memoirs about Socrates composed by Xenophon and Plato. Socrates, furthermore, was remembered by disciples who were longtime companions and eyewitnesses. Although the Gospels undoubtedly bear within them evidence of firsthand sources and even eyewitnesses, such material is not identified as such, and the narratives as a whole were most probably composed by authors of the generation after that of Jesus’ immediate followers.
"...Non narrative New Testament writings datable with some degree of probability before the year 70 testify to traditions circulating within the Christian movement concerning Jesus that correspond to important points within the Gospel narratives. Such traditions do not, by themselves, demonstrate historicity. But they demonstrate that memoires about Jesus were in fairly wide circulation. This makes it less likely that the corresponding points within the Gospels were the invention of a single author. If that were the case than such invention would have to be early enough and authoritative enough to have been distributed and unchallenged across the diverse communities with which Paul dealt. Such an hypothesis of course would work against the premise that Paul's form of Christianity had little to do with those shaping the memory of Jesus.".
42 And they devoted themselves to the apostles' teaching, to fellowship, to the breaking of bread, and to prayers. 43 Then fear came over everyone, and many wonders and signs were being performed through the apostles. 44 Now all the believers were together and had everything in common. 45 So they sold their possessions and property and distributed the proceeds to all, as anyone had a need. 46 And every day they devoted themselves to meeting together in the temple complex, and broke bread from house to house. They ate their food with gladness and simplicity of heart, 47 praising God and having favor with all the people. And every day the Lord added those being saved to them.
"No one is likely to deny that a tradition that is being handed on by word of mouth is likely to undergo modification. This is bound to happen, unless the tradition has been rigidly formulated and has been learned with careful safeguard against the intrusion of error" Tradition was controlled....[Neil adds in a fn:] "This is exactly the way in which the tradition was handed on among the Jews. IT is precisely on this ground that Scandinavian scholar H. Risenfeld in an essay entitled "The Gospel Tradition and its Beginnings" (1957) has passed some rather severe strictures on the form cuticle method. 
Communities that live in an oral culture tend to be story-telling communities. They sit around in long evenings telling and listening to stories--the same stories, over and over again. Such stories, especially when they are involved with memorable happenings that have determined in some way the existence and life of the particular group in question, acquire a fairly fixed form, down to precise phraseology (in narrative as well as in recorded speech), extremely early in their life--often within a day or so of the original incident taking place. They retain that form, and phraseology, as long as they are told. Each village and community has its recognized storytellers, the accredited bearers of its traditions; but the whole community knows the stories by heart, and if the teller varies them even slightly they will let him know in no uncertain terms. This matters quite a lot in cultures where, to this day, the desire to avoid 'shame' is a powerful motivation. 
If we knew that half of the information in a particular Gospel was based on “firsthand sources and even eyewitnesses”, then we might infer that at least half of the events or details in the Gospel were historically reliable (although without knowing anything about the personality, character, history, mental health and intelligence of the persons who were the supposed eyewitnesses, this would be a questionable inference), but since we don’t know which events or details have such backing, it would be the toss of a coin as to whether a given event or detail had such eyewitness evidence behind it. But we don’t even know this much. We don’t know whether 10% of the events and details of a particular Gospel are based on “firsthand sources and even eyewitnesses” or whether 30% or 50% or 70% of events and details are based on such evidence. Thus, the weak concession that Johnson makes here is of little significance.
As I have tried to show, the character of the Gospel narratives does not allow a fully satisfying reconstruction of Jesus ministry. Nevertheless certain fundamental points when taken together with confirming lines of convergence from outside testimony and non-narrative New Testament evidence, can be regarded as historical with a high degree of probability. Even the most critical historian can confidently assert that a Jew named Jesus worked as a teacher and wonder-worker in Palestine during the reign of Tiberius, was executed by crucifixion under the prefect Pontius Pilate, and continued to have followers after his death. These assertions are not mathematically or metaphysically certain, for certainty is not within the reach of history. But they enjoy a very high level of probability."
The Gospel of the Saviour, too. fits this description. Contrary' to popular opinion, Matthew, Mark, Luke and John were not included in the canon simply because they were the earliest gospels or because they were eyewitness accounts. Some non canonical gospels are dated roughly to the same period, and the canonical gospels and other early Christian accounts appear to rely on earlier reports. Thus, as far as the physical evidence is concerned, the canonical gospels do not take precedence over the noncanonical gospels (in terms of history--they do in doctrine). The fragments of John, Thomas and the Egerton Gospel share the distinction of being the earliest extant pieces of Christian writing known. And although the existing manuscript evidence for Thomas dates to the mid-second century, the scholars who first published the Greek fragments held open the possibility that it was actually composed in the first century, which would put it around the time John was composed.