You certainly can sequence your amplicons without cloning them. This will save time and money and of course a lot of effort. But you may miss out on some sequencing data on the 5' side. You can make up for that by sequencing from both 5' and 3' sides. This will also help you when identifying possible SNPs or other markers of interest.
The resulting electropherograms from direct sequencing may not be as clean as you would expect from a cloned product but I recon it is worth giving it a go.
I used to sequence viral coat protein genes (800 bp) with a forward and reverse primer combination and I also sequenced many bacterial 16S rRNA genes (1450 bp) using two forward and a single reverse primer. So range should not be too much of a problem.
Let me know if you get stuck or need some help in setting up the reactions.
It is acceptable to directly use PCR products for sequencing and then for use in phylogenetic analysis.
One piece of advice I will give, if you find what seem to be new haplotypes that differ from the 'norm' always re-amplify and re-sqeuence to avoid any PCR and sequencing errors.
If sample used for amplification was from single organism, you can directly perform sanger sequencing of amplicons. However if you have amplified a mixed sample (sample containing metagenomic, multiple organism DNA) in that case you need to perform cloning before sanger sequencing.
To avoid cloning steps, a new method of sequencing - Next Generation Sequencing is the best choice. Here you can use single or mixed sample source, amplify and sequence without cloning.
just to enlarge slightly on what Yashwant correctly says. Cloning isolates single sequences which may be unique. If you had a mixture of organisms and your sequenced pcr product had a 2 base polymorphism at 3 places in the sequenced product then there could be as many as 8 different distinguishable organisms present but the mixed sequence will not tell you which ones