I cannot remember ever having seen a graduate student present a PhD thesis proposal and be criticized for lack of ambition. It never happens. Even the weakest students—especially the weakest students—present proposals that are overly ambitious and that won’t ever get done, and certainly not in the 3-4 years remaining until graduation. In fact, in my experience it is exceedingly rare that a student presents a reasonable proposal, one that is actually doable during the remainder of the time in graduate school. Usually, those only happen when students “forget” to have their qualifying exams and end up presenting their “proposal” six months before the intended graduation date. In those cases, the students know that they won’t accomplish much new between proposal day and defense day, and basically present a proposal that consists entirely of completed work.
In Biology, most professors expect graduate students to complete about three projects, corresponding to the magical three specific aims in a typical grant proposal . In my mind, a graduate student who is defending her proposal, 2-3 years into her program, should have one project completed, one well under way, and one in the early planning stages. Students doing complicated experimental work might be less advanced at that time, but at a minimum they should have one project well under way by the time they defend their thesis proposal. This gives a pretty good rule of thumb for the amount of work the proposal should encompass: Aim 1 should be the work that is in the bag, and Aims 2 and 3 together should not require more than twice the amount of work already accomplished. Instead, what I commonly see is that the completed work is only a small component of Aim 1, which alone is going to take another two years to completion. Aim 2 will need four years on top of that, and Aim 3 another ten. Basically, many graduate students propose to carry out a lifetime of research during their graduate work.
I don’t quite know why PhD proposals tend to be overly ambitious. Maybe it’s youthful optimism or naiveté. I suspect, though, that there is a component of fear, the eternal graduate student fear of not being sufficiently productive, of not doing enough. Ironically, this fear often causes students to overlook the successes that are within reach and instead try to reach for the stars. In general, doing a successful thesis is a fine balancing act between being overly ambitious and playing it too safe, a topic for another post. However, there’s a difference between an actual PhD thesis and a thesis proposal: The thesis should contain some exciting, risky work, but for the proposal most professors want to see a plan that is doable, not one that might be doable if the stars align correctly. As a smart graduate student, you have two alternative plans, one safe and one daring, you work on both of them at the same time, and you present the safe one during the committee meeting.
A second, related issue I frequently notice is that students display poor judgment in how much work they can realistically accomplish in the remaining time. Estimates are consistently too optimistic. If you are in year three, and you have completed 50% of your first project, it is unlikely that you’ll complete this and two entirely different projects in the remaining 2-3 years of your PhD. Further, unless you’re a paper-writing machine, it’s unlikely that you can write a paper in less than three months. (And if you are a paper-writing machine, you should have plenty of papers by the time you defend your proposal anyway.) So, if you still have three manuscripts to complete, plus a thesis, the writing time alone is going to be about a year. If you’re already halfway into year three, you’ll have about another 18 months of actual research work you can do, because the other 12 you’ll spend writing. (Of course I’d recommend that you don’t wait all the way till the end before you start writing, but the math comes out the same.) My personal rule of thumb is things take about three times as long as students estimate things will take. So if a student says a particular project needs another three months in the lab plus a month to be written up, I expect that project to be done around the same time next year.
In conclusion, when you prepare your thesis proposal, realistically assess how much work you can complete during the remainder of your graduate years. Don't assume that your productivity will double or triple over the next two years, because it won’t. Budget at least three months for every paper you have to write, and triple the time you think it takes to complete the remaining lab work. If you have papers in review, consider that responding to reviewer comments and revising a paper frequently takes another two to three months, during which nothing else gets done. If you end up with a plan that will require another five years of work or more, then you’ll have to change your aims. See whether your current Aim 1 can be broken down into reasonable sub-aims which can be considered the separate chapters of your thesis. It’s quite common for me to conclude a PhD proposal defense by telling the student it'd be best to scrap Aims 2 and 3 altogether and instead expand Aim 1 into the entire thesis. If you come to that realization before the committee meeting, I don’t have to tell you so during, and everybody is happier.
 While proposals with either two or four aims can also be viable, two can appear as unimaginative (he really couldn’t think of anything else?) and four is getting dangerously close to being overly ambitious, so three it is.